"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
" and use a method that allows the use of a target. "
],
"text/plain": [
" and use a method that allows the use of a target."
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"Preparing data set. "
],
"text/plain": [
"Preparing data set."
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
" Data has 218621 observations with 6012 users and 3224 items. "
],
"text/plain": [
" Data has 218621 observations with 6012 users and 3224 items."
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
" Data prepared in: 0.237904s "
],
"text/plain": [
" Data prepared in: 0.237904s"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"218621 observations to process; with 3224 unique items. "
],
"text/plain": [
"218621 observations to process; with 3224 unique items."
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"pop = graphlab.popularity_recommender.create(train[train['rating'] > 4], 'user_id', 'movie_id')"
]
},
{
"cell_type": "code",
"execution_count": 12,
"metadata": {
"collapsed": false,
"slideshow": {
"slide_type": "slide"
}
},
"outputs": [
{
"data": {
"text/html": [
"Recsys training: model = ranking_factorization_recommender "
],
"text/plain": [
"Recsys training: model = ranking_factorization_recommender"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"Preparing data set. "
],
"text/plain": [
"Preparing data set."
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
" Data has 965508 observations with 6040 users and 3706 items. "
],
"text/plain": [
" Data has 965508 observations with 6040 users and 3706 items."
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
" Data prepared in: 0.907655s "
],
"text/plain": [
" Data prepared in: 0.907655s"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"Training ranking_factorization_recommender for recommendations. "
],
"text/plain": [
"Training ranking_factorization_recommender for recommendations."
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"+--------------------------------+--------------------------------------------------+----------+ "
],
"text/plain": [
"+--------------------------------+--------------------------------------------------+----------+"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| Parameter | Description | Value | "
],
"text/plain": [
"| Parameter | Description | Value |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"+--------------------------------+--------------------------------------------------+----------+ "
],
"text/plain": [
"+--------------------------------+--------------------------------------------------+----------+"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| num_factors | Factor Dimension | 32 | "
],
"text/plain": [
"| num_factors | Factor Dimension | 32 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| regularization | L2 Regularization on Factors | 1e-09 | "
],
"text/plain": [
"| regularization | L2 Regularization on Factors | 1e-09 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| solver | Solver used for training | adagrad | "
],
"text/plain": [
"| solver | Solver used for training | adagrad |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| linear_regularization | L2 Regularization on Linear Coefficients | 1e-09 | "
],
"text/plain": [
"| linear_regularization | L2 Regularization on Linear Coefficients | 1e-09 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| ranking_regularization | Rank-based Regularization Weight | 0.25 | "
],
"text/plain": [
"| ranking_regularization | Rank-based Regularization Weight | 0.25 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| max_iterations | Maximum Number of Iterations | 25 | "
],
"text/plain": [
"| max_iterations | Maximum Number of Iterations | 25 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"+--------------------------------+--------------------------------------------------+----------+ "
],
"text/plain": [
"+--------------------------------+--------------------------------------------------+----------+"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
" Optimizing model using SGD; tuning step size. "
],
"text/plain": [
" Optimizing model using SGD; tuning step size."
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
" Using 120688 / 965508 points for tuning the step size. "
],
"text/plain": [
" Using 120688 / 965508 points for tuning the step size."
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"+---------+-------------------+------------------------------------------+ "
],
"text/plain": [
"+---------+-------------------+------------------------------------------+"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| Attempt | Initial Step Size | Estimated Objective Value | "
],
"text/plain": [
"| Attempt | Initial Step Size | Estimated Objective Value |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"+---------+-------------------+------------------------------------------+ "
],
"text/plain": [
"+---------+-------------------+------------------------------------------+"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 0 | 16.6667 | Not Viable | "
],
"text/plain": [
"| 0 | 16.6667 | Not Viable |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 1 | 4.16667 | Not Viable | "
],
"text/plain": [
"| 1 | 4.16667 | Not Viable |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 2 | 1.04167 | Not Viable | "
],
"text/plain": [
"| 2 | 1.04167 | Not Viable |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 3 | 0.260417 | 1.80479 | "
],
"text/plain": [
"| 3 | 0.260417 | 1.80479 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 4 | 0.130208 | 1.8322 | "
],
"text/plain": [
"| 4 | 0.130208 | 1.8322 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 5 | 0.0651042 | 1.8873 | "
],
"text/plain": [
"| 5 | 0.0651042 | 1.8873 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 6 | 0.0325521 | 1.88706 | "
],
"text/plain": [
"| 6 | 0.0325521 | 1.88706 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"+---------+-------------------+------------------------------------------+ "
],
"text/plain": [
"+---------+-------------------+------------------------------------------+"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| Final | 0.260417 | 1.80479 | "
],
"text/plain": [
"| Final | 0.260417 | 1.80479 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"+---------+-------------------+------------------------------------------+ "
],
"text/plain": [
"+---------+-------------------+------------------------------------------+"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"Starting Optimization. "
],
"text/plain": [
"Starting Optimization."
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"+---------+--------------+-------------------+-----------------------+-------------+ "
],
"text/plain": [
"+---------+--------------+-------------------+-----------------------+-------------+"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| Iter. | Elapsed Time | Approx. Objective | Approx. Training RMSE | Step Size | "
],
"text/plain": [
"| Iter. | Elapsed Time | Approx. Objective | Approx. Training RMSE | Step Size |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"+---------+--------------+-------------------+-----------------------+-------------+ "
],
"text/plain": [
"+---------+--------------+-------------------+-----------------------+-------------+"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| Initial | 101us | 2.4462 | 1.11698 | | "
],
"text/plain": [
"| Initial | 101us | 2.4462 | 1.11698 | |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"+---------+--------------+-------------------+-----------------------+-------------+ "
],
"text/plain": [
"+---------+--------------+-------------------+-----------------------+-------------+"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 1 | 1.43s | DIVERGED | DIVERGED | 0.260417 | "
],
"text/plain": [
"| 1 | 1.43s | DIVERGED | DIVERGED | 0.260417 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| RESET | 1.91s | 2.44619 | 1.11697 | | "
],
"text/plain": [
"| RESET | 1.91s | 2.44619 | 1.11697 | |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 1 | 3.24s | DIVERGED | DIVERGED | 0.130208 | "
],
"text/plain": [
"| 1 | 3.24s | DIVERGED | DIVERGED | 0.130208 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| RESET | 3.75s | 2.44619 | 1.11697 | | "
],
"text/plain": [
"| RESET | 3.75s | 2.44619 | 1.11697 | |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 1 | 4.84s | 2.10443 | 1.14093 | 0.0651042 | "
],
"text/plain": [
"| 1 | 4.84s | 2.10443 | 1.14093 | 0.0651042 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 2 | 5.81s | 1.82027 | 1.04353 | 0.0651042 | "
],
"text/plain": [
"| 2 | 5.81s | 1.82027 | 1.04353 | 0.0651042 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 3 | 6.89s | 1.75645 | 1.02196 | 0.0651042 | "
],
"text/plain": [
"| 3 | 6.89s | 1.75645 | 1.02196 | 0.0651042 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 4 | 7.90s | 1.7206 | 1.01294 | 0.0651042 | "
],
"text/plain": [
"| 4 | 7.90s | 1.7206 | 1.01294 | 0.0651042 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 5 | 8.99s | 1.69207 | 1.00488 | 0.0651042 | "
],
"text/plain": [
"| 5 | 8.99s | 1.69207 | 1.00488 | 0.0651042 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 6 | 10.03s | 1.66916 | 0.998471 | 0.0651042 | "
],
"text/plain": [
"| 6 | 10.03s | 1.66916 | 0.998471 | 0.0651042 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 7 | 11.11s | 1.64975 | 0.992687 | 0.0651042 | "
],
"text/plain": [
"| 7 | 11.11s | 1.64975 | 0.992687 | 0.0651042 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 8 | 12.27s | 1.63331 | 0.987803 | 0.0651042 | "
],
"text/plain": [
"| 8 | 12.27s | 1.63331 | 0.987803 | 0.0651042 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 9 | 13.53s | 1.6203 | 0.984347 | 0.0651042 | "
],
"text/plain": [
"| 9 | 13.53s | 1.6203 | 0.984347 | 0.0651042 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 10 | 14.68s | 1.60869 | 0.981751 | 0.0651042 | "
],
"text/plain": [
"| 10 | 14.68s | 1.60869 | 0.981751 | 0.0651042 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 11 | 15.79s | 1.59758 | 0.977906 | 0.0651042 | "
],
"text/plain": [
"| 11 | 15.79s | 1.59758 | 0.977906 | 0.0651042 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 12 | 16.82s | 1.58984 | 0.976171 | 0.0651042 | "
],
"text/plain": [
"| 12 | 16.82s | 1.58984 | 0.976171 | 0.0651042 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 13 | 17.96s | 1.58036 | 0.973489 | 0.0651042 | "
],
"text/plain": [
"| 13 | 17.96s | 1.58036 | 0.973489 | 0.0651042 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 14 | 19.24s | 1.57243 | 0.971477 | 0.0651042 | "
],
"text/plain": [
"| 14 | 19.24s | 1.57243 | 0.971477 | 0.0651042 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 15 | 20.37s | 1.56503 | 0.969302 | 0.0651042 | "
],
"text/plain": [
"| 15 | 20.37s | 1.56503 | 0.969302 | 0.0651042 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 16 | 21.50s | 1.55807 | 0.967444 | 0.0651042 | "
],
"text/plain": [
"| 16 | 21.50s | 1.55807 | 0.967444 | 0.0651042 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 17 | 22.63s | 1.55118 | 0.965764 | 0.0651042 | "
],
"text/plain": [
"| 17 | 22.63s | 1.55118 | 0.965764 | 0.0651042 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 18 | 23.79s | 1.54509 | 0.963793 | 0.0651042 | "
],
"text/plain": [
"| 18 | 23.79s | 1.54509 | 0.963793 | 0.0651042 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 19 | 24.93s | 1.53942 | 0.961991 | 0.0651042 | "
],
"text/plain": [
"| 19 | 24.93s | 1.53942 | 0.961991 | 0.0651042 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 20 | 26.23s | 1.53433 | 0.960398 | 0.0651042 | "
],
"text/plain": [
"| 20 | 26.23s | 1.53433 | 0.960398 | 0.0651042 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 21 | 27.37s | 1.52844 | 0.959103 | 0.0651042 | "
],
"text/plain": [
"| 21 | 27.37s | 1.52844 | 0.959103 | 0.0651042 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 22 | 28.46s | 1.52382 | 0.958025 | 0.0651042 | "
],
"text/plain": [
"| 22 | 28.46s | 1.52382 | 0.958025 | 0.0651042 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 23 | 29.55s | 1.51829 | 0.956181 | 0.0651042 | "
],
"text/plain": [
"| 23 | 29.55s | 1.51829 | 0.956181 | 0.0651042 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 24 | 30.81s | 1.51352 | 0.955045 | 0.0651042 | "
],
"text/plain": [
"| 24 | 30.81s | 1.51352 | 0.955045 | 0.0651042 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 25 | 32.13s | 1.50902 | 0.953533 | 0.0651042 | "
],
"text/plain": [
"| 25 | 32.13s | 1.50902 | 0.953533 | 0.0651042 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"+---------+--------------+-------------------+-----------------------+-------------+ "
],
"text/plain": [
"+---------+--------------+-------------------+-----------------------+-------------+"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"Optimization Complete: Maximum number of passes through the data reached. "
],
"text/plain": [
"Optimization Complete: Maximum number of passes through the data reached."
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"Computing final objective value and training RMSE. "
],
"text/plain": [
"Computing final objective value and training RMSE."
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
" Final objective value: 1.53876 "
],
"text/plain": [
" Final objective value: 1.53876"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
" Final training RMSE: 0.948071 "
],
"text/plain": [
" Final training RMSE: 0.948071"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"m = graphlab.recommender.create(train, 'user_id', 'movie_id', 'rating')"
]
},
{
"cell_type": "code",
"execution_count": 13,
"metadata": {
"collapsed": false,
"slideshow": {
"slide_type": "slide"
}
},
"outputs": [
{
"data": {
"text/plain": [
"Class : RankingFactorizationRecommender\n",
"\n",
"Schema\n",
"------\n",
"User ID : user_id\n",
"Item ID : movie_id\n",
"Target : rating\n",
"Additional observation features : 1\n",
"Number of user side features : 0\n",
"Number of item side features : 0\n",
"\n",
"Statistics\n",
"----------\n",
"Number of observations : 965508\n",
"Number of users : 6040\n",
"Number of items : 3706\n",
"\n",
"Training summary\n",
"----------------\n",
"Training time : 36.9965\n",
"\n",
"Model Parameters\n",
"----------------\n",
"Model class : RankingFactorizationRecommender\n",
"num_factors : 32\n",
"binary_target : 0\n",
"side_data_factorization : 1\n",
"solver : auto\n",
"nmf : 0\n",
"max_iterations : 25\n",
"\n",
"Regularization Settings\n",
"-----------------------\n",
"regularization : 0.0\n",
"regularization_type : normal\n",
"linear_regularization : 0.0\n",
"ranking_regularization : 0.25\n",
"unobserved_rating_value : -1.79769313486e+308\n",
"num_sampled_negative_examples : 4\n",
"ials_confidence_scaling_type : auto\n",
"ials_confidence_scaling_factor : 1\n",
"\n",
"Optimization Settings\n",
"---------------------\n",
"init_random_sigma : 0.01\n",
"sgd_convergence_interval : 4\n",
"sgd_convergence_threshold : 0.0\n",
"sgd_max_trial_iterations : 5\n",
"sgd_sampling_block_size : 131072\n",
"sgd_step_adjustment_interval : 4\n",
"sgd_step_size : 0.0\n",
"sgd_trial_sample_minimum_size : 10000\n",
"sgd_trial_sample_proportion : 0.125\n",
"step_size_decrease_rate : 0.75\n",
"additional_iterations_if_unhealthy: 5\n",
"adagrad_momentum_weighting : 0.9\n",
"num_tempering_iterations : 4\n",
"tempering_regularization_start_value: 0.0\n",
"track_exact_loss : 0"
]
},
"execution_count": 13,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"m"
]
},
{
"cell_type": "code",
"execution_count": 14,
"metadata": {
"collapsed": false,
"slideshow": {
"slide_type": "slide"
}
},
"outputs": [
{
"data": {
"text/plain": [
"{'intercept': 3.5821495005738013, 'movie_id': Columns:\n",
" \tmovie_id\tint\n",
" \tlinear_terms\tfloat\n",
" \tfactors\tarray\n",
" \n",
" Rows: 3706\n",
" \n",
" Data:\n",
" +----------+------------------+-------------------------------+\n",
" | movie_id | linear_terms | factors |\n",
" +----------+------------------+-------------------------------+\n",
" | 1193 | 1.06781125069 | [-0.119829073548, -0.02245... |\n",
" | 661 | -0.0261590108275 | [-0.727257788181, 0.016146... |\n",
" | 914 | 0.324085891247 | [-0.859803378582, 0.056376... |\n",
" | 3408 | 0.565778970718 | [0.334619760513, -0.014206... |\n",
" | 2355 | 0.648248255253 | [-0.248598009348, 0.103843... |\n",
" | 1197 | 1.12024652958 | [-0.100379563868, 0.085359... |\n",
" | 1287 | 0.345532894135 | [-0.247123196721, 0.024613... |\n",
" | 2804 | 0.894821941853 | [-0.272583067417, 0.046351... |\n",
" | 594 | 0.311594575644 | [-0.974369823933, 0.054282... |\n",
" | 919 | 0.97704321146 | [-0.598346889019, 0.085630... |\n",
" +----------+------------------+-------------------------------+\n",
" [3706 rows x 3 columns]\n",
" Note: Only the head of the SFrame is printed.\n",
" You can use print_rows(num_rows=m, num_columns=n) to print more rows and columns., 'side_data': Columns:\n",
" \tfeature\tstr\n",
" \tindex\tstr\n",
" \tlinear_terms\tfloat\n",
" \tfactors\tarray\n",
" \n",
" Rows: 1\n",
" \n",
" Data:\n",
" +-----------+-------+-----------------+-------------------------------+\n",
" | feature | index | linear_terms | factors |\n",
" +-----------+-------+-----------------+-------------------------------+\n",
" | timestamp | 0 | -0.116745471954 | [-0.564183712006, 1.267165... |\n",
" +-----------+-------+-----------------+-------------------------------+\n",
" [1 rows x 4 columns], 'user_id': Columns:\n",
" \tuser_id\tint\n",
" \tlinear_terms\tfloat\n",
" \tfactors\tarray\n",
" \n",
" Rows: 6040\n",
" \n",
" Data:\n",
" +---------+------------------+-------------------------------+\n",
" | user_id | linear_terms | factors |\n",
" +---------+------------------+-------------------------------+\n",
" | 1 | -0.027785371989 | [-0.0942558199167, 0.00739... |\n",
" | 2 | -0.0234720371664 | [0.015922004357, -0.033992... |\n",
" | 3 | -0.0345229320228 | [0.176564618945, -0.050576... |\n",
" | 4 | -0.0198582224548 | [-0.0773911848664, -0.0500... |\n",
" | 5 | -0.0562275871634 | [-0.0598151274025, -0.0059... |\n",
" | 6 | -0.0401206016541 | [0.0565584115684, 0.030123... |\n",
" | 7 | -0.0433877147734 | [0.205288589001, -0.060017... |\n",
" | 8 | -0.0184100158513 | [0.169030055404, -0.043373... |\n",
" | 9 | -0.0512112490833 | [0.163330376148, -0.060946... |\n",
" | 10 | -0.0407416447997 | [-0.420519113541, 0.110337... |\n",
" +---------+------------------+-------------------------------+\n",
" [6040 rows x 3 columns]\n",
" Note: Only the head of the SFrame is printed.\n",
" You can use print_rows(num_rows=m, num_columns=n) to print more rows and columns.}"
]
},
"execution_count": 14,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"m['coefficients']"
]
},
{
"cell_type": "code",
"execution_count": 16,
"metadata": {
"collapsed": false,
"slideshow": {
"slide_type": "slide"
}
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"compare_models: using 183 users to estimate model performance\n",
"PROGRESS: Evaluate model M0\n",
"\n",
"Precision and recall summary statistics by cutoff\n",
"+--------+-----------------+-----------------+\n",
"| cutoff | mean_precision | mean_recall |\n",
"+--------+-----------------+-----------------+\n",
"| 1 | 0.109289617486 | 0.0154169412131 |\n",
"| 2 | 0.114754098361 | 0.0315571827129 |\n",
"| 3 | 0.103825136612 | 0.0393550677194 |\n",
"| 4 | 0.0983606557377 | 0.0488860172488 |\n",
"| 5 | 0.0983606557377 | 0.057530354299 |\n",
"| 6 | 0.0983606557377 | 0.06952814808 |\n",
"| 7 | 0.0967993754879 | 0.0776744105871 |\n",
"| 8 | 0.0949453551913 | 0.0871933441083 |\n",
"| 9 | 0.0910746812386 | 0.0970583805009 |\n",
"| 10 | 0.0890710382514 | 0.105731522781 |\n",
"+--------+-----------------+-----------------+\n",
"[10 rows x 3 columns]\n",
"\n",
"PROGRESS: Evaluate model M1\n",
"\n",
"Precision and recall summary statistics by cutoff\n",
"+--------+----------------+-----------------+\n",
"| cutoff | mean_precision | mean_recall |\n",
"+--------+----------------+-----------------+\n",
"| 1 | 0.218579234973 | 0.0336356996991 |\n",
"| 2 | 0.185792349727 | 0.0491081612808 |\n",
"| 3 | 0.182149362477 | 0.0721856847862 |\n",
"| 4 | 0.172131147541 | 0.086814432767 |\n",
"| 5 | 0.165027322404 | 0.0983099175722 |\n",
"| 6 | 0.152094717668 | 0.110996252299 |\n",
"| 7 | 0.148321623731 | 0.13067735829 |\n",
"| 8 | 0.148907103825 | 0.150453968213 |\n",
"| 9 | 0.142076502732 | 0.158699171088 |\n",
"| 10 | 0.134426229508 | 0.166857542043 |\n",
"+--------+----------------+-----------------+\n",
"[10 rows x 3 columns]\n",
"\n",
"PROGRESS: Evaluate model M2\n",
"\n",
"Precision and recall summary statistics by cutoff\n",
"+--------+----------------+-----------------+\n",
"| cutoff | mean_precision | mean_recall |\n",
"+--------+----------------+-----------------+\n",
"| 1 | 0.27868852459 | 0.0355923139267 |\n",
"| 2 | 0.226775956284 | 0.0540712203094 |\n",
"| 3 | 0.213114754098 | 0.0716753913564 |\n",
"| 4 | 0.198087431694 | 0.0898091945474 |\n",
"| 5 | 0.183606557377 | 0.100699809919 |\n",
"| 6 | 0.182149362477 | 0.11362028645 |\n",
"| 7 | 0.185011709602 | 0.137198290932 |\n",
"| 8 | 0.177595628415 | 0.147966304582 |\n",
"| 9 | 0.169398907104 | 0.156916738229 |\n",
"| 10 | 0.160655737705 | 0.171367047623 |\n",
"+--------+----------------+-----------------+\n",
"[10 rows x 3 columns]\n",
"\n"
]
},
{
"data": {
"text/plain": [
"[{'precision_recall_by_user': Columns:\n",
" \tuser_id\tint\n",
" \tcutoff\tint\n",
" \tprecision\tfloat\n",
" \trecall\tfloat\n",
" \tcount\tint\n",
" \n",
" Rows: 3294\n",
" \n",
" Data:\n",
" +---------+--------+----------------+----------------+-------+\n",
" | user_id | cutoff | precision | recall | count |\n",
" +---------+--------+----------------+----------------+-------+\n",
" | 42 | 1 | 0.0 | 0.0 | 7 |\n",
" | 42 | 2 | 0.5 | 0.142857142857 | 7 |\n",
" | 42 | 3 | 0.333333333333 | 0.142857142857 | 7 |\n",
" | 42 | 4 | 0.25 | 0.142857142857 | 7 |\n",
" | 42 | 5 | 0.2 | 0.142857142857 | 7 |\n",
" | 42 | 6 | 0.166666666667 | 0.142857142857 | 7 |\n",
" | 42 | 7 | 0.142857142857 | 0.142857142857 | 7 |\n",
" | 42 | 8 | 0.125 | 0.142857142857 | 7 |\n",
" | 42 | 9 | 0.111111111111 | 0.142857142857 | 7 |\n",
" | 42 | 10 | 0.1 | 0.142857142857 | 7 |\n",
" +---------+--------+----------------+----------------+-------+\n",
" [3294 rows x 5 columns]\n",
" Note: Only the head of the SFrame is printed.\n",
" You can use print_rows(num_rows=m, num_columns=n) to print more rows and columns.,\n",
" 'precision_recall_overall': Columns:\n",
" \tcutoff\tint\n",
" \tprecision\tfloat\n",
" \trecall\tfloat\n",
" \n",
" Rows: 18\n",
" \n",
" Data:\n",
" +--------+-----------------+-----------------+\n",
" | cutoff | precision | recall |\n",
" +--------+-----------------+-----------------+\n",
" | 1 | 0.109289617486 | 0.0154169412131 |\n",
" | 2 | 0.114754098361 | 0.0315571827129 |\n",
" | 3 | 0.103825136612 | 0.0393550677194 |\n",
" | 4 | 0.0983606557377 | 0.0488860172488 |\n",
" | 5 | 0.0983606557377 | 0.057530354299 |\n",
" | 6 | 0.0983606557377 | 0.06952814808 |\n",
" | 7 | 0.0967993754879 | 0.0776744105871 |\n",
" | 8 | 0.0949453551913 | 0.0871933441083 |\n",
" | 9 | 0.0910746812386 | 0.0970583805009 |\n",
" | 10 | 0.0890710382514 | 0.105731522781 |\n",
" +--------+-----------------+-----------------+\n",
" [18 rows x 3 columns]\n",
" Note: Only the head of the SFrame is printed.\n",
" You can use print_rows(num_rows=m, num_columns=n) to print more rows and columns.},\n",
" {'precision_recall_by_user': Columns:\n",
" \tuser_id\tint\n",
" \tcutoff\tint\n",
" \tprecision\tfloat\n",
" \trecall\tfloat\n",
" \tcount\tint\n",
" \n",
" Rows: 3294\n",
" \n",
" Data:\n",
" +---------+--------+----------------+----------------+-------+\n",
" | user_id | cutoff | precision | recall | count |\n",
" +---------+--------+----------------+----------------+-------+\n",
" | 42 | 1 | 1.0 | 0.142857142857 | 7 |\n",
" | 42 | 2 | 0.5 | 0.142857142857 | 7 |\n",
" | 42 | 3 | 0.333333333333 | 0.142857142857 | 7 |\n",
" | 42 | 4 | 0.25 | 0.142857142857 | 7 |\n",
" | 42 | 5 | 0.4 | 0.285714285714 | 7 |\n",
" | 42 | 6 | 0.333333333333 | 0.285714285714 | 7 |\n",
" | 42 | 7 | 0.285714285714 | 0.285714285714 | 7 |\n",
" | 42 | 8 | 0.25 | 0.285714285714 | 7 |\n",
" | 42 | 9 | 0.222222222222 | 0.285714285714 | 7 |\n",
" | 42 | 10 | 0.2 | 0.285714285714 | 7 |\n",
" +---------+--------+----------------+----------------+-------+\n",
" [3294 rows x 5 columns]\n",
" Note: Only the head of the SFrame is printed.\n",
" You can use print_rows(num_rows=m, num_columns=n) to print more rows and columns.,\n",
" 'precision_recall_overall': Columns:\n",
" \tcutoff\tint\n",
" \tprecision\tfloat\n",
" \trecall\tfloat\n",
" \n",
" Rows: 18\n",
" \n",
" Data:\n",
" +--------+----------------+-----------------+\n",
" | cutoff | precision | recall |\n",
" +--------+----------------+-----------------+\n",
" | 1 | 0.218579234973 | 0.0336356996991 |\n",
" | 2 | 0.185792349727 | 0.0491081612808 |\n",
" | 3 | 0.182149362477 | 0.0721856847862 |\n",
" | 4 | 0.172131147541 | 0.086814432767 |\n",
" | 5 | 0.165027322404 | 0.0983099175722 |\n",
" | 6 | 0.152094717668 | 0.110996252299 |\n",
" | 7 | 0.148321623731 | 0.13067735829 |\n",
" | 8 | 0.148907103825 | 0.150453968213 |\n",
" | 9 | 0.142076502732 | 0.158699171088 |\n",
" | 10 | 0.134426229508 | 0.166857542043 |\n",
" +--------+----------------+-----------------+\n",
" [18 rows x 3 columns]\n",
" Note: Only the head of the SFrame is printed.\n",
" You can use print_rows(num_rows=m, num_columns=n) to print more rows and columns.},\n",
" {'precision_recall_by_user': Columns:\n",
" \tuser_id\tint\n",
" \tcutoff\tint\n",
" \tprecision\tfloat\n",
" \trecall\tfloat\n",
" \tcount\tint\n",
" \n",
" Rows: 3294\n",
" \n",
" Data:\n",
" +---------+--------+----------------+----------------+-------+\n",
" | user_id | cutoff | precision | recall | count |\n",
" +---------+--------+----------------+----------------+-------+\n",
" | 42 | 1 | 1.0 | 0.142857142857 | 7 |\n",
" | 42 | 2 | 0.5 | 0.142857142857 | 7 |\n",
" | 42 | 3 | 0.333333333333 | 0.142857142857 | 7 |\n",
" | 42 | 4 | 0.25 | 0.142857142857 | 7 |\n",
" | 42 | 5 | 0.2 | 0.142857142857 | 7 |\n",
" | 42 | 6 | 0.166666666667 | 0.142857142857 | 7 |\n",
" | 42 | 7 | 0.142857142857 | 0.142857142857 | 7 |\n",
" | 42 | 8 | 0.125 | 0.142857142857 | 7 |\n",
" | 42 | 9 | 0.111111111111 | 0.142857142857 | 7 |\n",
" | 42 | 10 | 0.2 | 0.285714285714 | 7 |\n",
" +---------+--------+----------------+----------------+-------+\n",
" [3294 rows x 5 columns]\n",
" Note: Only the head of the SFrame is printed.\n",
" You can use print_rows(num_rows=m, num_columns=n) to print more rows and columns.,\n",
" 'precision_recall_overall': Columns:\n",
" \tcutoff\tint\n",
" \tprecision\tfloat\n",
" \trecall\tfloat\n",
" \n",
" Rows: 18\n",
" \n",
" Data:\n",
" +--------+----------------+-----------------+\n",
" | cutoff | precision | recall |\n",
" +--------+----------------+-----------------+\n",
" | 1 | 0.27868852459 | 0.0355923139267 |\n",
" | 2 | 0.226775956284 | 0.0540712203094 |\n",
" | 3 | 0.213114754098 | 0.0716753913564 |\n",
" | 4 | 0.198087431694 | 0.0898091945474 |\n",
" | 5 | 0.183606557377 | 0.100699809919 |\n",
" | 6 | 0.182149362477 | 0.11362028645 |\n",
" | 7 | 0.185011709602 | 0.137198290932 |\n",
" | 8 | 0.177595628415 | 0.147966304582 |\n",
" | 9 | 0.169398907104 | 0.156916738229 |\n",
" | 10 | 0.160655737705 | 0.171367047623 |\n",
" +--------+----------------+-----------------+\n",
" [18 rows x 3 columns]\n",
" Note: Only the head of the SFrame is printed.\n",
" You can use print_rows(num_rows=m, num_columns=n) to print more rows and columns.}]"
]
},
"execution_count": 16,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"graphlab.recommender.util.compare_models(test[test['rating'] > 4], \n",
" [pop, itemcf, m], \n",
" user_sample=0.2, \n",
" metric='precision_recall')"
]
},
{
"cell_type": "markdown",
"metadata": {
"slideshow": {
"slide_type": "slide"
}
},
"source": [
"### Optimizing for ranking"
]
},
{
"cell_type": "code",
"execution_count": 20,
"metadata": {
"collapsed": false,
"slideshow": {
"slide_type": "slide"
}
},
"outputs": [
{
"data": {
"text/html": [
"Recsys training: model = ranking_factorization_recommender "
],
"text/plain": [
"Recsys training: model = ranking_factorization_recommender"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"Preparing data set. "
],
"text/plain": [
"Preparing data set."
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
" Data has 965508 observations with 6040 users and 3706 items. "
],
"text/plain": [
" Data has 965508 observations with 6040 users and 3706 items."
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
" Data prepared in: 0.910656s "
],
"text/plain": [
" Data prepared in: 0.910656s"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"Training ranking_factorization_recommender for recommendations. "
],
"text/plain": [
"Training ranking_factorization_recommender for recommendations."
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"+--------------------------------+--------------------------------------------------+----------+ "
],
"text/plain": [
"+--------------------------------+--------------------------------------------------+----------+"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| Parameter | Description | Value | "
],
"text/plain": [
"| Parameter | Description | Value |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"+--------------------------------+--------------------------------------------------+----------+ "
],
"text/plain": [
"+--------------------------------+--------------------------------------------------+----------+"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| num_factors | Factor Dimension | 32 | "
],
"text/plain": [
"| num_factors | Factor Dimension | 32 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| regularization | L2 Regularization on Factors | 1e-09 | "
],
"text/plain": [
"| regularization | L2 Regularization on Factors | 1e-09 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| solver | Solver used for training | adagrad | "
],
"text/plain": [
"| solver | Solver used for training | adagrad |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| linear_regularization | L2 Regularization on Linear Coefficients | 1e-09 | "
],
"text/plain": [
"| linear_regularization | L2 Regularization on Linear Coefficients | 1e-09 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| ranking_regularization | Rank-based Regularization Weight | 0.25 | "
],
"text/plain": [
"| ranking_regularization | Rank-based Regularization Weight | 0.25 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| unobserved_rating_value | Ranking Target Rating for Unobserved Interacti...| 3 | "
],
"text/plain": [
"| unobserved_rating_value | Ranking Target Rating for Unobserved Interacti...| 3 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| max_iterations | Maximum Number of Iterations | 25 | "
],
"text/plain": [
"| max_iterations | Maximum Number of Iterations | 25 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"+--------------------------------+--------------------------------------------------+----------+ "
],
"text/plain": [
"+--------------------------------+--------------------------------------------------+----------+"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
" Optimizing model using SGD; tuning step size. "
],
"text/plain": [
" Optimizing model using SGD; tuning step size."
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
" Using 120688 / 965508 points for tuning the step size. "
],
"text/plain": [
" Using 120688 / 965508 points for tuning the step size."
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"+---------+-------------------+------------------------------------------+ "
],
"text/plain": [
"+---------+-------------------+------------------------------------------+"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| Attempt | Initial Step Size | Estimated Objective Value | "
],
"text/plain": [
"| Attempt | Initial Step Size | Estimated Objective Value |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"+---------+-------------------+------------------------------------------+ "
],
"text/plain": [
"+---------+-------------------+------------------------------------------+"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 0 | 16.6667 | Not Viable | "
],
"text/plain": [
"| 0 | 16.6667 | Not Viable |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 1 | 4.16667 | Not Viable | "
],
"text/plain": [
"| 1 | 4.16667 | Not Viable |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 2 | 1.04167 | Not Viable | "
],
"text/plain": [
"| 2 | 1.04167 | Not Viable |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 3 | 0.260417 | Not Viable | "
],
"text/plain": [
"| 3 | 0.260417 | Not Viable |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 4 | 0.0651042 | 0.998804 | "
],
"text/plain": [
"| 4 | 0.0651042 | 0.998804 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 5 | 0.0325521 | 0.953073 | "
],
"text/plain": [
"| 5 | 0.0325521 | 0.953073 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 6 | 0.016276 | 1.00856 | "
],
"text/plain": [
"| 6 | 0.016276 | 1.00856 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 7 | 0.00813802 | 1.0661 | "
],
"text/plain": [
"| 7 | 0.00813802 | 1.0661 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 8 | 0.00406901 | 1.23488 | "
],
"text/plain": [
"| 8 | 0.00406901 | 1.23488 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"+---------+-------------------+------------------------------------------+ "
],
"text/plain": [
"+---------+-------------------+------------------------------------------+"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| Final | 0.0325521 | 0.953073 | "
],
"text/plain": [
"| Final | 0.0325521 | 0.953073 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"+---------+-------------------+------------------------------------------+ "
],
"text/plain": [
"+---------+-------------------+------------------------------------------+"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"Starting Optimization. "
],
"text/plain": [
"Starting Optimization."
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"+---------+--------------+-------------------+-----------------------+-------------+ "
],
"text/plain": [
"+---------+--------------+-------------------+-----------------------+-------------+"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| Iter. | Elapsed Time | Approx. Objective | Approx. Training RMSE | Step Size | "
],
"text/plain": [
"| Iter. | Elapsed Time | Approx. Objective | Approx. Training RMSE | Step Size |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"+---------+--------------+-------------------+-----------------------+-------------+ "
],
"text/plain": [
"+---------+--------------+-------------------+-----------------------+-------------+"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| Initial | 107us | 1.33247 | 1.11699 | | "
],
"text/plain": [
"| Initial | 107us | 1.33247 | 1.11699 | |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"+---------+--------------+-------------------+-----------------------+-------------+ "
],
"text/plain": [
"+---------+--------------+-------------------+-----------------------+-------------+"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 1 | 885.253ms | 1.10069 | 1.00012 | 0.0325521 | "
],
"text/plain": [
"| 1 | 885.253ms | 1.10069 | 1.00012 | 0.0325521 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 2 | 1.90s | 1.01103 | 0.956979 | 0.0325521 | "
],
"text/plain": [
"| 2 | 1.90s | 1.01103 | 0.956979 | 0.0325521 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 3 | 3.00s | 0.974312 | 0.937847 | 0.0325521 | "
],
"text/plain": [
"| 3 | 3.00s | 0.974312 | 0.937847 | 0.0325521 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 4 | 4.07s | 0.960712 | 0.931004 | 0.0325521 | "
],
"text/plain": [
"| 4 | 4.07s | 0.960712 | 0.931004 | 0.0325521 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 5 | 5.12s | 0.949761 | 0.9254 | 0.0325521 | "
],
"text/plain": [
"| 5 | 5.12s | 0.949761 | 0.9254 | 0.0325521 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 6 | 6.07s | 0.942225 | 0.921526 | 0.0325521 | "
],
"text/plain": [
"| 6 | 6.07s | 0.942225 | 0.921526 | 0.0325521 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 7 | 7.06s | 0.935704 | 0.918205 | 0.0325521 | "
],
"text/plain": [
"| 7 | 7.06s | 0.935704 | 0.918205 | 0.0325521 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 8 | 8.02s | 0.930567 | 0.915684 | 0.0325521 | "
],
"text/plain": [
"| 8 | 8.02s | 0.930567 | 0.915684 | 0.0325521 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 9 | 9.05s | 0.925405 | 0.912967 | 0.0325521 | "
],
"text/plain": [
"| 9 | 9.05s | 0.925405 | 0.912967 | 0.0325521 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 10 | 10.33s | 0.920952 | 0.910727 | 0.0325521 | "
],
"text/plain": [
"| 10 | 10.33s | 0.920952 | 0.910727 | 0.0325521 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 11 | 11.58s | 0.916647 | 0.908743 | 0.0325521 | "
],
"text/plain": [
"| 11 | 11.58s | 0.916647 | 0.908743 | 0.0325521 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 12 | 12.62s | 0.913399 | 0.907017 | 0.0325521 | "
],
"text/plain": [
"| 12 | 12.62s | 0.913399 | 0.907017 | 0.0325521 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 13 | 13.58s | 0.909575 | 0.904969 | 0.0325521 | "
],
"text/plain": [
"| 13 | 13.58s | 0.909575 | 0.904969 | 0.0325521 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 14 | 14.67s | 0.906824 | 0.90367 | 0.0325521 | "
],
"text/plain": [
"| 14 | 14.67s | 0.906824 | 0.90367 | 0.0325521 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 15 | 15.77s | 0.904054 | 0.902198 | 0.0325521 | "
],
"text/plain": [
"| 15 | 15.77s | 0.904054 | 0.902198 | 0.0325521 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 16 | 16.86s | 0.901294 | 0.90096 | 0.0325521 | "
],
"text/plain": [
"| 16 | 16.86s | 0.901294 | 0.90096 | 0.0325521 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 17 | 17.85s | 0.898579 | 0.899525 | 0.0325521 | "
],
"text/plain": [
"| 17 | 17.85s | 0.898579 | 0.899525 | 0.0325521 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 18 | 18.86s | 0.896474 | 0.898482 | 0.0325521 | "
],
"text/plain": [
"| 18 | 18.86s | 0.896474 | 0.898482 | 0.0325521 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 19 | 20.07s | 0.894312 | 0.897331 | 0.0325521 | "
],
"text/plain": [
"| 19 | 20.07s | 0.894312 | 0.897331 | 0.0325521 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 20 | 21.10s | 0.892068 | 0.896046 | 0.0325521 | "
],
"text/plain": [
"| 20 | 21.10s | 0.892068 | 0.896046 | 0.0325521 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 21 | 22.12s | 0.88988 | 0.894963 | 0.0325521 | "
],
"text/plain": [
"| 21 | 22.12s | 0.88988 | 0.894963 | 0.0325521 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 22 | 23.33s | 0.887669 | 0.893956 | 0.0325521 | "
],
"text/plain": [
"| 22 | 23.33s | 0.887669 | 0.893956 | 0.0325521 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 23 | 24.51s | 0.885674 | 0.892851 | 0.0325521 | "
],
"text/plain": [
"| 23 | 24.51s | 0.885674 | 0.892851 | 0.0325521 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 24 | 25.58s | 0.884228 | 0.892176 | 0.0325521 | "
],
"text/plain": [
"| 24 | 25.58s | 0.884228 | 0.892176 | 0.0325521 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 25 | 26.59s | 0.882557 | 0.891299 | 0.0325521 | "
],
"text/plain": [
"| 25 | 26.59s | 0.882557 | 0.891299 | 0.0325521 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"+---------+--------------+-------------------+-----------------------+-------------+ "
],
"text/plain": [
"+---------+--------------+-------------------+-----------------------+-------------+"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"Optimization Complete: Maximum number of passes through the data reached. "
],
"text/plain": [
"Optimization Complete: Maximum number of passes through the data reached."
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"Computing final objective value and training RMSE. "
],
"text/plain": [
"Computing final objective value and training RMSE."
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
" Final objective value: 0.882406 "
],
"text/plain": [
" Final objective value: 0.882406"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
" Final training RMSE: 0.886832 "
],
"text/plain": [
" Final training RMSE: 0.886832"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"m_rank = graphlab.recommender.ranking_factorization_recommender.create(train, 'user_id', 'movie_id', 'rating', \n",
" unobserved_rating_value=3)"
]
},
{
"cell_type": "code",
"execution_count": 21,
"metadata": {
"collapsed": false,
"slideshow": {
"slide_type": "slide"
}
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"compare_models: using 183 users to estimate model performance\n",
"PROGRESS: Evaluate model M0\n",
"\n",
"Precision and recall summary statistics by cutoff\n",
"+--------+-----------------+-----------------+\n",
"| cutoff | mean_precision | mean_recall |\n",
"+--------+-----------------+-----------------+\n",
"| 1 | 0.103825136612 | 0.0116730652166 |\n",
"| 2 | 0.101092896175 | 0.0282821265133 |\n",
"| 3 | 0.0819672131148 | 0.0387227556041 |\n",
"| 4 | 0.0833333333333 | 0.054260741295 |\n",
"| 5 | 0.0786885245902 | 0.0652009814042 |\n",
"| 6 | 0.0765027322404 | 0.0758533021658 |\n",
"| 7 | 0.0772833723653 | 0.087870157321 |\n",
"| 8 | 0.0785519125683 | 0.10363872715 |\n",
"| 9 | 0.0740740740741 | 0.108125993351 |\n",
"| 10 | 0.0743169398907 | 0.125169969412 |\n",
"+--------+-----------------+-----------------+\n",
"[10 rows x 3 columns]\n",
"\n",
"PROGRESS: Evaluate model M1\n",
"\n",
"Precision and recall summary statistics by cutoff\n",
"+--------+----------------+-----------------+\n",
"| cutoff | mean_precision | mean_recall |\n",
"+--------+----------------+-----------------+\n",
"| 1 | 0.16393442623 | 0.033918559494 |\n",
"| 2 | 0.166666666667 | 0.0559035280067 |\n",
"| 3 | 0.158469945355 | 0.0802610557096 |\n",
"| 4 | 0.147540983607 | 0.0993935937697 |\n",
"| 5 | 0.138797814208 | 0.116006405262 |\n",
"| 6 | 0.128415300546 | 0.125824450712 |\n",
"| 7 | 0.128805620609 | 0.148368313836 |\n",
"| 8 | 0.125 | 0.162294248876 |\n",
"| 9 | 0.12204007286 | 0.173015991344 |\n",
"| 10 | 0.117486338798 | 0.18606052953 |\n",
"+--------+----------------+-----------------+\n",
"[10 rows x 3 columns]\n",
"\n",
"PROGRESS: Evaluate model M2\n",
"\n",
"Precision and recall summary statistics by cutoff\n",
"+--------+----------------+-----------------+\n",
"| cutoff | mean_precision | mean_recall |\n",
"+--------+----------------+-----------------+\n",
"| 1 | 0.24043715847 | 0.032990686752 |\n",
"| 2 | 0.196721311475 | 0.0593716586723 |\n",
"| 3 | 0.187613843352 | 0.0783908312908 |\n",
"| 4 | 0.180327868852 | 0.102234190817 |\n",
"| 5 | 0.16393442623 | 0.120610724355 |\n",
"| 6 | 0.159380692168 | 0.140119556509 |\n",
"| 7 | 0.152224824356 | 0.157806327365 |\n",
"| 8 | 0.142759562842 | 0.166809497193 |\n",
"| 9 | 0.137826350941 | 0.175280850522 |\n",
"| 10 | 0.134972677596 | 0.186140806231 |\n",
"+--------+----------------+-----------------+\n",
"[10 rows x 3 columns]\n",
"\n",
"PROGRESS: Evaluate model M3\n",
"\n",
"Precision and recall summary statistics by cutoff\n",
"+--------+----------------+-----------------+\n",
"| cutoff | mean_precision | mean_recall |\n",
"+--------+----------------+-----------------+\n",
"| 1 | 0.114754098361 | 0.0220448212961 |\n",
"| 2 | 0.120218579235 | 0.0351999242985 |\n",
"| 3 | 0.111111111111 | 0.0445415535372 |\n",
"| 4 | 0.106557377049 | 0.052975921608 |\n",
"| 5 | 0.110382513661 | 0.076614241225 |\n",
"| 6 | 0.111111111111 | 0.0956858236539 |\n",
"| 7 | 0.110850897736 | 0.113159656127 |\n",
"| 8 | 0.106557377049 | 0.121421484491 |\n",
"| 9 | 0.102610807529 | 0.136444329342 |\n",
"| 10 | 0.101092896175 | 0.14669712098 |\n",
"+--------+----------------+-----------------+\n",
"[10 rows x 3 columns]\n",
"\n"
]
}
],
"source": [
"results = graphlab.recommender.util.compare_models(test[test['rating'] > 4], \n",
" [pop, itemcf, m, m_rank], \n",
" user_sample=0.2, \n",
" metric='precision_recall')"
]
},
{
"cell_type": "code",
"execution_count": 28,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/html": [
"\n",
" \n",
" cutoff \n",
" precision \n",
" recall \n",
" \n",
" \n",
" 1 \n",
" 0.114754098361 \n",
" 0.0220448212961 \n",
" \n",
" \n",
" 2 \n",
" 0.120218579235 \n",
" 0.0351999242985 \n",
" \n",
" \n",
" 3 \n",
" 0.111111111111 \n",
" 0.0445415535372 \n",
" \n",
" \n",
" 4 \n",
" 0.106557377049 \n",
" 0.052975921608 \n",
" \n",
" \n",
" 5 \n",
" 0.110382513661 \n",
" 0.076614241225 \n",
" \n",
" \n",
" 6 \n",
" 0.111111111111 \n",
" 0.0956858236539 \n",
" \n",
" \n",
" 7 \n",
" 0.110850897736 \n",
" 0.113159656127 \n",
" \n",
" \n",
" 8 \n",
" 0.106557377049 \n",
" 0.121421484491 \n",
" \n",
" \n",
" 9 \n",
" 0.102610807529 \n",
" 0.136444329342 \n",
" \n",
" \n",
" 10 \n",
" 0.101092896175 \n",
" 0.14669712098 \n",
" \n",
"
\n",
"[18 rows x 3 columns]
Note: Only the head of the SFrame is printed.
You can use print_rows(num_rows=m, num_columns=n) to print more rows and columns.\n",
"
Recsys training: model = ranking_factorization_recommender "
],
"text/plain": [
"Recsys training: model = ranking_factorization_recommender"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"Preparing data set. "
],
"text/plain": [
"Preparing data set."
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
" Data has 965508 observations with 6040 users and 3706 items. "
],
"text/plain": [
" Data has 965508 observations with 6040 users and 3706 items."
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
" Data prepared in: 0.872319s "
],
"text/plain": [
" Data prepared in: 0.872319s"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"Training ranking_factorization_recommender for recommendations. "
],
"text/plain": [
"Training ranking_factorization_recommender for recommendations."
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"+--------------------------------+--------------------------------------------------+----------+ "
],
"text/plain": [
"+--------------------------------+--------------------------------------------------+----------+"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| Parameter | Description | Value | "
],
"text/plain": [
"| Parameter | Description | Value |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"+--------------------------------+--------------------------------------------------+----------+ "
],
"text/plain": [
"+--------------------------------+--------------------------------------------------+----------+"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| num_factors | Factor Dimension | 32 | "
],
"text/plain": [
"| num_factors | Factor Dimension | 32 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| regularization | L2 Regularization on Factors | 1e-09 | "
],
"text/plain": [
"| regularization | L2 Regularization on Factors | 1e-09 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| solver | Solver used for training | adagrad | "
],
"text/plain": [
"| solver | Solver used for training | adagrad |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| linear_regularization | L2 Regularization on Linear Coefficients | 1e-09 | "
],
"text/plain": [
"| linear_regularization | L2 Regularization on Linear Coefficients | 1e-09 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| ranking_regularization | Rank-based Regularization Weight | 0.25 | "
],
"text/plain": [
"| ranking_regularization | Rank-based Regularization Weight | 0.25 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| side_data_factorization | Assign Factors for Side Data | True | "
],
"text/plain": [
"| side_data_factorization | Assign Factors for Side Data | True |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| max_iterations | Maximum Number of Iterations | 25 | "
],
"text/plain": [
"| max_iterations | Maximum Number of Iterations | 25 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"+--------------------------------+--------------------------------------------------+----------+ "
],
"text/plain": [
"+--------------------------------+--------------------------------------------------+----------+"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
" Optimizing model using SGD; tuning step size. "
],
"text/plain": [
" Optimizing model using SGD; tuning step size."
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
" Using 120688 / 965508 points for tuning the step size. "
],
"text/plain": [
" Using 120688 / 965508 points for tuning the step size."
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"+---------+-------------------+------------------------------------------+ "
],
"text/plain": [
"+---------+-------------------+------------------------------------------+"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| Attempt | Initial Step Size | Estimated Objective Value | "
],
"text/plain": [
"| Attempt | Initial Step Size | Estimated Objective Value |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"+---------+-------------------+------------------------------------------+ "
],
"text/plain": [
"+---------+-------------------+------------------------------------------+"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 0 | 7.14286 | Not Viable | "
],
"text/plain": [
"| 0 | 7.14286 | Not Viable |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 1 | 1.78571 | Not Viable | "
],
"text/plain": [
"| 1 | 1.78571 | Not Viable |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 2 | 0.446429 | 1.52387 | "
],
"text/plain": [
"| 2 | 0.446429 | 1.52387 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 3 | 0.223214 | Not Viable | "
],
"text/plain": [
"| 3 | 0.223214 | Not Viable |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 4 | 0.0558036 | 1.79945 | "
],
"text/plain": [
"| 4 | 0.0558036 | 1.79945 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 5 | 0.0279018 | 1.78969 | "
],
"text/plain": [
"| 5 | 0.0279018 | 1.78969 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"+---------+-------------------+------------------------------------------+ "
],
"text/plain": [
"+---------+-------------------+------------------------------------------+"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| Final | 0.446429 | 1.52387 | "
],
"text/plain": [
"| Final | 0.446429 | 1.52387 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"+---------+-------------------+------------------------------------------+ "
],
"text/plain": [
"+---------+-------------------+------------------------------------------+"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"Starting Optimization. "
],
"text/plain": [
"Starting Optimization."
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"+---------+--------------+-------------------+-----------------------+-------------+ "
],
"text/plain": [
"+---------+--------------+-------------------+-----------------------+-------------+"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| Iter. | Elapsed Time | Approx. Objective | Approx. Training RMSE | Step Size | "
],
"text/plain": [
"| Iter. | Elapsed Time | Approx. Objective | Approx. Training RMSE | Step Size |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"+---------+--------------+-------------------+-----------------------+-------------+ "
],
"text/plain": [
"+---------+--------------+-------------------+-----------------------+-------------+"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| Initial | 63us | 2.44592 | 1.11697 | | "
],
"text/plain": [
"| Initial | 63us | 2.44592 | 1.11697 | |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"+---------+--------------+-------------------+-----------------------+-------------+ "
],
"text/plain": [
"+---------+--------------+-------------------+-----------------------+-------------+"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 1 | 2.13s | DIVERGED | DIVERGED | 0.446429 | "
],
"text/plain": [
"| 1 | 2.13s | DIVERGED | DIVERGED | 0.446429 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| RESET | 2.99s | 2.44647 | 1.11698 | | "
],
"text/plain": [
"| RESET | 2.99s | 2.44647 | 1.11698 | |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 1 | 5.10s | 1.83529 | 1.10141 | 0.223214 | "
],
"text/plain": [
"| 1 | 5.10s | 1.83529 | 1.10141 | 0.223214 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 2 | 6.90s | 1.47109 | 0.95455 | 0.223214 | "
],
"text/plain": [
"| 2 | 6.90s | 1.47109 | 0.95455 | 0.223214 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 3 | 8.80s | 1.36577 | 0.917865 | 0.223214 | "
],
"text/plain": [
"| 3 | 8.80s | 1.36577 | 0.917865 | 0.223214 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 4 | 10.62s | 1.31007 | 0.897384 | 0.223214 | "
],
"text/plain": [
"| 4 | 10.62s | 1.31007 | 0.897384 | 0.223214 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 5 | 12.44s | 1.27125 | 0.883491 | 0.223214 | "
],
"text/plain": [
"| 5 | 12.44s | 1.27125 | 0.883491 | 0.223214 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 6 | 14.67s | 1.24635 | 0.873351 | 0.223214 | "
],
"text/plain": [
"| 6 | 14.67s | 1.24635 | 0.873351 | 0.223214 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 7 | 16.70s | 1.22622 | 0.865748 | 0.223214 | "
],
"text/plain": [
"| 7 | 16.70s | 1.22622 | 0.865748 | 0.223214 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 8 | 18.63s | 1.21028 | 0.859754 | 0.223214 | "
],
"text/plain": [
"| 8 | 18.63s | 1.21028 | 0.859754 | 0.223214 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 9 | 20.50s | 1.19796 | 0.854523 | 0.223214 | "
],
"text/plain": [
"| 9 | 20.50s | 1.19796 | 0.854523 | 0.223214 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 10 | 22.46s | 1.18689 | 0.850478 | 0.223214 | "
],
"text/plain": [
"| 10 | 22.46s | 1.18689 | 0.850478 | 0.223214 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 11 | 24.31s | 1.1783 | 0.846785 | 0.223214 | "
],
"text/plain": [
"| 11 | 24.31s | 1.1783 | 0.846785 | 0.223214 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 12 | 26.18s | 1.17084 | 0.84352 | 0.223214 | "
],
"text/plain": [
"| 12 | 26.18s | 1.17084 | 0.84352 | 0.223214 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 13 | 28.07s | 1.16355 | 0.840745 | 0.223214 | "
],
"text/plain": [
"| 13 | 28.07s | 1.16355 | 0.840745 | 0.223214 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 14 | 29.97s | 1.15711 | 0.838398 | 0.223214 | "
],
"text/plain": [
"| 14 | 29.97s | 1.15711 | 0.838398 | 0.223214 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 15 | 32.09s | 1.15247 | 0.836416 | 0.223214 | "
],
"text/plain": [
"| 15 | 32.09s | 1.15247 | 0.836416 | 0.223214 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 16 | 34.04s | 1.14785 | 0.83443 | 0.223214 | "
],
"text/plain": [
"| 16 | 34.04s | 1.14785 | 0.83443 | 0.223214 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 17 | 36.25s | 1.14331 | 0.832546 | 0.223214 | "
],
"text/plain": [
"| 17 | 36.25s | 1.14331 | 0.832546 | 0.223214 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 18 | 38.41s | 1.13848 | 0.830724 | 0.223214 | "
],
"text/plain": [
"| 18 | 38.41s | 1.13848 | 0.830724 | 0.223214 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 19 | 40.34s | 1.13683 | 0.82959 | 0.223214 | "
],
"text/plain": [
"| 19 | 40.34s | 1.13683 | 0.82959 | 0.223214 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 20 | 42.39s | 1.13266 | 0.828052 | 0.223214 | "
],
"text/plain": [
"| 20 | 42.39s | 1.13266 | 0.828052 | 0.223214 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 21 | 44.75s | 1.13008 | 0.827049 | 0.223214 | "
],
"text/plain": [
"| 21 | 44.75s | 1.13008 | 0.827049 | 0.223214 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 22 | 47.05s | 1.12695 | 0.82589 | 0.223214 | "
],
"text/plain": [
"| 22 | 47.05s | 1.12695 | 0.82589 | 0.223214 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 23 | 49.58s | 1.12374 | 0.824564 | 0.223214 | "
],
"text/plain": [
"| 23 | 49.58s | 1.12374 | 0.824564 | 0.223214 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 24 | 51.93s | 1.12202 | 0.823962 | 0.223214 | "
],
"text/plain": [
"| 24 | 51.93s | 1.12202 | 0.823962 | 0.223214 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 25 | 54.30s | 1.12015 | 0.822855 | 0.223214 | "
],
"text/plain": [
"| 25 | 54.30s | 1.12015 | 0.822855 | 0.223214 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"+---------+--------------+-------------------+-----------------------+-------------+ "
],
"text/plain": [
"+---------+--------------+-------------------+-----------------------+-------------+"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"Optimization Complete: Maximum number of passes through the data reached. "
],
"text/plain": [
"Optimization Complete: Maximum number of passes through the data reached."
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"Computing final objective value and training RMSE. "
],
"text/plain": [
"Computing final objective value and training RMSE."
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
" Final objective value: 1.12099 "
],
"text/plain": [
" Final objective value: 1.12099"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
" Final training RMSE: 0.797943 "
],
"text/plain": [
" Final training RMSE: 0.797943"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"Recsys training: model = ranking_factorization_recommender "
],
"text/plain": [
"Recsys training: model = ranking_factorization_recommender"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"Preparing data set. "
],
"text/plain": [
"Preparing data set."
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
" Data has 965508 observations with 6040 users and 3883 items. "
],
"text/plain": [
" Data has 965508 observations with 6040 users and 3883 items."
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
" Data prepared in: 1.16364s "
],
"text/plain": [
" Data prepared in: 1.16364s"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"Training ranking_factorization_recommender for recommendations. "
],
"text/plain": [
"Training ranking_factorization_recommender for recommendations."
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"+--------------------------------+--------------------------------------------------+----------+ "
],
"text/plain": [
"+--------------------------------+--------------------------------------------------+----------+"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| Parameter | Description | Value | "
],
"text/plain": [
"| Parameter | Description | Value |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"+--------------------------------+--------------------------------------------------+----------+ "
],
"text/plain": [
"+--------------------------------+--------------------------------------------------+----------+"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| num_factors | Factor Dimension | 32 | "
],
"text/plain": [
"| num_factors | Factor Dimension | 32 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| regularization | L2 Regularization on Factors | 1e-09 | "
],
"text/plain": [
"| regularization | L2 Regularization on Factors | 1e-09 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| solver | Solver used for training | adagrad | "
],
"text/plain": [
"| solver | Solver used for training | adagrad |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| linear_regularization | L2 Regularization on Linear Coefficients | 1e-09 | "
],
"text/plain": [
"| linear_regularization | L2 Regularization on Linear Coefficients | 1e-09 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| ranking_regularization | Rank-based Regularization Weight | 0.25 | "
],
"text/plain": [
"| ranking_regularization | Rank-based Regularization Weight | 0.25 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| side_data_factorization | Assign Factors for Side Data | True | "
],
"text/plain": [
"| side_data_factorization | Assign Factors for Side Data | True |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| max_iterations | Maximum Number of Iterations | 25 | "
],
"text/plain": [
"| max_iterations | Maximum Number of Iterations | 25 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"+--------------------------------+--------------------------------------------------+----------+ "
],
"text/plain": [
"+--------------------------------+--------------------------------------------------+----------+"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
" Optimizing model using SGD; tuning step size. "
],
"text/plain": [
" Optimizing model using SGD; tuning step size."
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
" Using 120688 / 965508 points for tuning the step size. "
],
"text/plain": [
" Using 120688 / 965508 points for tuning the step size."
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"+---------+-------------------+------------------------------------------+ "
],
"text/plain": [
"+---------+-------------------+------------------------------------------+"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| Attempt | Initial Step Size | Estimated Objective Value | "
],
"text/plain": [
"| Attempt | Initial Step Size | Estimated Objective Value |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"+---------+-------------------+------------------------------------------+ "
],
"text/plain": [
"+---------+-------------------+------------------------------------------+"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 0 | 10 | Not Viable | "
],
"text/plain": [
"| 0 | 10 | Not Viable |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 1 | 2.5 | Not Viable | "
],
"text/plain": [
"| 1 | 2.5 | Not Viable |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 2 | 0.625 | Not Viable | "
],
"text/plain": [
"| 2 | 0.625 | Not Viable |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 3 | 0.15625 | 1.08443 | "
],
"text/plain": [
"| 3 | 0.15625 | 1.08443 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 4 | 0.078125 | 1.72974 | "
],
"text/plain": [
"| 4 | 0.078125 | 1.72974 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 5 | 0.0390625 | 1.84654 | "
],
"text/plain": [
"| 5 | 0.0390625 | 1.84654 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 6 | 0.0195312 | 1.74856 | "
],
"text/plain": [
"| 6 | 0.0195312 | 1.74856 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"+---------+-------------------+------------------------------------------+ "
],
"text/plain": [
"+---------+-------------------+------------------------------------------+"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| Final | 0.15625 | 1.08443 | "
],
"text/plain": [
"| Final | 0.15625 | 1.08443 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"+---------+-------------------+------------------------------------------+ "
],
"text/plain": [
"+---------+-------------------+------------------------------------------+"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"Starting Optimization. "
],
"text/plain": [
"Starting Optimization."
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"+---------+--------------+-------------------+-----------------------+-------------+ "
],
"text/plain": [
"+---------+--------------+-------------------+-----------------------+-------------+"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| Iter. | Elapsed Time | Approx. Objective | Approx. Training RMSE | Step Size | "
],
"text/plain": [
"| Iter. | Elapsed Time | Approx. Objective | Approx. Training RMSE | Step Size |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"+---------+--------------+-------------------+-----------------------+-------------+ "
],
"text/plain": [
"+---------+--------------+-------------------+-----------------------+-------------+"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| Initial | 70us | 2.44637 | 1.11698 | | "
],
"text/plain": [
"| Initial | 70us | 2.44637 | 1.11698 | |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"+---------+--------------+-------------------+-----------------------+-------------+ "
],
"text/plain": [
"+---------+--------------+-------------------+-----------------------+-------------+"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 1 | 2.33s | DIVERGED | DIVERGED | 0.15625 | "
],
"text/plain": [
"| 1 | 2.33s | DIVERGED | DIVERGED | 0.15625 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| RESET | 3.11s | 2.44644 | 1.11697 | | "
],
"text/plain": [
"| RESET | 3.11s | 2.44644 | 1.11697 | |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 1 | 4.92s | 1.75622 | 1.06161 | 0.078125 | "
],
"text/plain": [
"| 1 | 4.92s | 1.75622 | 1.06161 | 0.078125 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 2 | 6.57s | 1.50614 | 0.966049 | 0.078125 | "
],
"text/plain": [
"| 2 | 6.57s | 1.50614 | 0.966049 | 0.078125 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 3 | 8.08s | 1.4167 | 0.933697 | 0.078125 | "
],
"text/plain": [
"| 3 | 8.08s | 1.4167 | 0.933697 | 0.078125 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 4 | 9.53s | 1.37077 | 0.917417 | 0.078125 | "
],
"text/plain": [
"| 4 | 9.53s | 1.37077 | 0.917417 | 0.078125 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 5 | 10.99s | 1.34379 | 0.908108 | 0.078125 | "
],
"text/plain": [
"| 5 | 10.99s | 1.34379 | 0.908108 | 0.078125 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 6 | 12.65s | 1.32151 | 0.899986 | 0.078125 | "
],
"text/plain": [
"| 6 | 12.65s | 1.32151 | 0.899986 | 0.078125 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 7 | 14.14s | 1.30427 | 0.89374 | 0.078125 | "
],
"text/plain": [
"| 7 | 14.14s | 1.30427 | 0.89374 | 0.078125 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 8 | 15.59s | 1.2895 | 0.888268 | 0.078125 | "
],
"text/plain": [
"| 8 | 15.59s | 1.2895 | 0.888268 | 0.078125 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 9 | 17.03s | 1.27727 | 0.884087 | 0.078125 | "
],
"text/plain": [
"| 9 | 17.03s | 1.27727 | 0.884087 | 0.078125 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 10 | 18.47s | 1.2662 | 0.879697 | 0.078125 | "
],
"text/plain": [
"| 10 | 18.47s | 1.2662 | 0.879697 | 0.078125 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 11 | 19.91s | 1.25785 | 0.876548 | 0.078125 | "
],
"text/plain": [
"| 11 | 19.91s | 1.25785 | 0.876548 | 0.078125 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 12 | 21.33s | 1.24893 | 0.873322 | 0.078125 | "
],
"text/plain": [
"| 12 | 21.33s | 1.24893 | 0.873322 | 0.078125 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 13 | 22.73s | 1.24222 | 0.870908 | 0.078125 | "
],
"text/plain": [
"| 13 | 22.73s | 1.24222 | 0.870908 | 0.078125 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 14 | 24.17s | 1.23693 | 0.868724 | 0.078125 | "
],
"text/plain": [
"| 14 | 24.17s | 1.23693 | 0.868724 | 0.078125 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 15 | 25.64s | 1.23104 | 0.866697 | 0.078125 | "
],
"text/plain": [
"| 15 | 25.64s | 1.23104 | 0.866697 | 0.078125 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 16 | 27.07s | 1.22657 | 0.865066 | 0.078125 | "
],
"text/plain": [
"| 16 | 27.07s | 1.22657 | 0.865066 | 0.078125 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 17 | 28.47s | 1.22185 | 0.86311 | 0.078125 | "
],
"text/plain": [
"| 17 | 28.47s | 1.22185 | 0.86311 | 0.078125 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 18 | 29.90s | 1.21603 | 0.860832 | 0.078125 | "
],
"text/plain": [
"| 18 | 29.90s | 1.21603 | 0.860832 | 0.078125 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 19 | 31.34s | 1.21214 | 0.859841 | 0.078125 | "
],
"text/plain": [
"| 19 | 31.34s | 1.21214 | 0.859841 | 0.078125 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 20 | 32.75s | 1.20866 | 0.858349 | 0.078125 | "
],
"text/plain": [
"| 20 | 32.75s | 1.20866 | 0.858349 | 0.078125 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 21 | 34.17s | 1.20588 | 0.857265 | 0.078125 | "
],
"text/plain": [
"| 21 | 34.17s | 1.20588 | 0.857265 | 0.078125 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 22 | 35.59s | 1.2013 | 0.855384 | 0.078125 | "
],
"text/plain": [
"| 22 | 35.59s | 1.2013 | 0.855384 | 0.078125 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 23 | 36.98s | 1.19868 | 0.854415 | 0.078125 | "
],
"text/plain": [
"| 23 | 36.98s | 1.19868 | 0.854415 | 0.078125 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 24 | 38.40s | 1.19618 | 0.8536 | 0.078125 | "
],
"text/plain": [
"| 24 | 38.40s | 1.19618 | 0.8536 | 0.078125 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 25 | 39.82s | 1.19373 | 0.852524 | 0.078125 | "
],
"text/plain": [
"| 25 | 39.82s | 1.19373 | 0.852524 | 0.078125 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"+---------+--------------+-------------------+-----------------------+-------------+ "
],
"text/plain": [
"+---------+--------------+-------------------+-----------------------+-------------+"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"Optimization Complete: Maximum number of passes through the data reached. "
],
"text/plain": [
"Optimization Complete: Maximum number of passes through the data reached."
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"Computing final objective value and training RMSE. "
],
"text/plain": [
"Computing final objective value and training RMSE."
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
" Final objective value: 1.22299 "
],
"text/plain": [
" Final objective value: 1.22299"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
" Final training RMSE: 0.842974 "
],
"text/plain": [
" Final training RMSE: 0.842974"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"Recsys training: model = ranking_factorization_recommender "
],
"text/plain": [
"Recsys training: model = ranking_factorization_recommender"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"Preparing data set. "
],
"text/plain": [
"Preparing data set."
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
" Data has 965508 observations with 6040 users and 3883 items. "
],
"text/plain": [
" Data has 965508 observations with 6040 users and 3883 items."
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
" Data prepared in: 0.897359s "
],
"text/plain": [
" Data prepared in: 0.897359s"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"Training ranking_factorization_recommender for recommendations. "
],
"text/plain": [
"Training ranking_factorization_recommender for recommendations."
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"+--------------------------------+--------------------------------------------------+----------+ "
],
"text/plain": [
"+--------------------------------+--------------------------------------------------+----------+"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| Parameter | Description | Value | "
],
"text/plain": [
"| Parameter | Description | Value |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"+--------------------------------+--------------------------------------------------+----------+ "
],
"text/plain": [
"+--------------------------------+--------------------------------------------------+----------+"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| num_factors | Factor Dimension | 32 | "
],
"text/plain": [
"| num_factors | Factor Dimension | 32 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| regularization | L2 Regularization on Factors | 1e-09 | "
],
"text/plain": [
"| regularization | L2 Regularization on Factors | 1e-09 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| solver | Solver used for training | adagrad | "
],
"text/plain": [
"| solver | Solver used for training | adagrad |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| linear_regularization | L2 Regularization on Linear Coefficients | 1e-09 | "
],
"text/plain": [
"| linear_regularization | L2 Regularization on Linear Coefficients | 1e-09 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| ranking_regularization | Rank-based Regularization Weight | 0.25 | "
],
"text/plain": [
"| ranking_regularization | Rank-based Regularization Weight | 0.25 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| side_data_factorization | Assign Factors for Side Data | True | "
],
"text/plain": [
"| side_data_factorization | Assign Factors for Side Data | True |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| max_iterations | Maximum Number of Iterations | 25 | "
],
"text/plain": [
"| max_iterations | Maximum Number of Iterations | 25 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"+--------------------------------+--------------------------------------------------+----------+ "
],
"text/plain": [
"+--------------------------------+--------------------------------------------------+----------+"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
" Optimizing model using SGD; tuning step size. "
],
"text/plain": [
" Optimizing model using SGD; tuning step size."
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
" Using 120688 / 965508 points for tuning the step size. "
],
"text/plain": [
" Using 120688 / 965508 points for tuning the step size."
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"+---------+-------------------+------------------------------------------+ "
],
"text/plain": [
"+---------+-------------------+------------------------------------------+"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| Attempt | Initial Step Size | Estimated Objective Value | "
],
"text/plain": [
"| Attempt | Initial Step Size | Estimated Objective Value |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"+---------+-------------------+------------------------------------------+ "
],
"text/plain": [
"+---------+-------------------+------------------------------------------+"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 0 | 5.55556 | Not Viable | "
],
"text/plain": [
"| 0 | 5.55556 | Not Viable |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 1 | 1.38889 | Not Viable | "
],
"text/plain": [
"| 1 | 1.38889 | Not Viable |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 2 | 0.347222 | 1.41163 | "
],
"text/plain": [
"| 2 | 0.347222 | 1.41163 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 3 | 0.173611 | 1.06475 | "
],
"text/plain": [
"| 3 | 0.173611 | 1.06475 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 4 | 0.0868056 | 1.03146 | "
],
"text/plain": [
"| 4 | 0.0868056 | 1.03146 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 5 | 0.0434028 | 1.1885 | "
],
"text/plain": [
"| 5 | 0.0434028 | 1.1885 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 6 | 0.0217014 | 1.54004 | "
],
"text/plain": [
"| 6 | 0.0217014 | 1.54004 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 7 | 0.0108507 | 1.77557 | "
],
"text/plain": [
"| 7 | 0.0108507 | 1.77557 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"+---------+-------------------+------------------------------------------+ "
],
"text/plain": [
"+---------+-------------------+------------------------------------------+"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| Final | 0.0868056 | 1.03146 | "
],
"text/plain": [
"| Final | 0.0868056 | 1.03146 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"+---------+-------------------+------------------------------------------+ "
],
"text/plain": [
"+---------+-------------------+------------------------------------------+"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"Starting Optimization. "
],
"text/plain": [
"Starting Optimization."
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"+---------+--------------+-------------------+-----------------------+-------------+ "
],
"text/plain": [
"+---------+--------------+-------------------+-----------------------+-------------+"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| Iter. | Elapsed Time | Approx. Objective | Approx. Training RMSE | Step Size | "
],
"text/plain": [
"| Iter. | Elapsed Time | Approx. Objective | Approx. Training RMSE | Step Size |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"+---------+--------------+-------------------+-----------------------+-------------+ "
],
"text/plain": [
"+---------+--------------+-------------------+-----------------------+-------------+"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| Initial | 70us | 2.4467 | 1.117 | | "
],
"text/plain": [
"| Initial | 70us | 2.4467 | 1.117 | |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"+---------+--------------+-------------------+-----------------------+-------------+ "
],
"text/plain": [
"+---------+--------------+-------------------+-----------------------+-------------+"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 1 | 3.02s | DIVERGED | DIVERGED | 0.0868056 | "
],
"text/plain": [
"| 1 | 3.02s | DIVERGED | DIVERGED | 0.0868056 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| RESET | 3.97s | 2.4468 | 1.11698 | | "
],
"text/plain": [
"| RESET | 3.97s | 2.4468 | 1.11698 | |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 1 | 6.22s | 1.56747 | 1.00211 | 0.0434028 | "
],
"text/plain": [
"| 1 | 6.22s | 1.56747 | 1.00211 | 0.0434028 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 2 | 8.40s | 1.4123 | 0.936364 | 0.0434028 | "
],
"text/plain": [
"| 2 | 8.40s | 1.4123 | 0.936364 | 0.0434028 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 3 | 10.61s | 1.34465 | 0.91251 | 0.0434028 | "
],
"text/plain": [
"| 3 | 10.61s | 1.34465 | 0.91251 | 0.0434028 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 4 | 13.00s | 1.30528 | 0.898029 | 0.0434028 | "
],
"text/plain": [
"| 4 | 13.00s | 1.30528 | 0.898029 | 0.0434028 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 5 | 15.21s | 1.27777 | 0.887646 | 0.0434028 | "
],
"text/plain": [
"| 5 | 15.21s | 1.27777 | 0.887646 | 0.0434028 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 6 | 18.09s | 1.25523 | 0.879139 | 0.0434028 | "
],
"text/plain": [
"| 6 | 18.09s | 1.25523 | 0.879139 | 0.0434028 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 7 | 20.54s | 1.23908 | 0.872961 | 0.0434028 | "
],
"text/plain": [
"| 7 | 20.54s | 1.23908 | 0.872961 | 0.0434028 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 8 | 22.79s | 1.22569 | 0.867368 | 0.0434028 | "
],
"text/plain": [
"| 8 | 22.79s | 1.22569 | 0.867368 | 0.0434028 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 9 | 25.25s | 1.21337 | 0.862705 | 0.0434028 | "
],
"text/plain": [
"| 9 | 25.25s | 1.21337 | 0.862705 | 0.0434028 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 10 | 28.35s | 1.20338 | 0.858906 | 0.0434028 | "
],
"text/plain": [
"| 10 | 28.35s | 1.20338 | 0.858906 | 0.0434028 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 11 | 31.26s | 1.19501 | 0.855631 | 0.0434028 | "
],
"text/plain": [
"| 11 | 31.26s | 1.19501 | 0.855631 | 0.0434028 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 12 | 33.70s | 1.18678 | 0.852674 | 0.0434028 | "
],
"text/plain": [
"| 12 | 33.70s | 1.18678 | 0.852674 | 0.0434028 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 13 | 35.86s | 1.18128 | 0.850198 | 0.0434028 | "
],
"text/plain": [
"| 13 | 35.86s | 1.18128 | 0.850198 | 0.0434028 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 14 | 38.05s | 1.17569 | 0.847772 | 0.0434028 | "
],
"text/plain": [
"| 14 | 38.05s | 1.17569 | 0.847772 | 0.0434028 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 15 | 40.25s | 1.16953 | 0.845739 | 0.0434028 | "
],
"text/plain": [
"| 15 | 40.25s | 1.16953 | 0.845739 | 0.0434028 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 16 | 42.45s | 1.16479 | 0.843867 | 0.0434028 | "
],
"text/plain": [
"| 16 | 42.45s | 1.16479 | 0.843867 | 0.0434028 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 17 | 44.62s | 1.16132 | 0.842166 | 0.0434028 | "
],
"text/plain": [
"| 17 | 44.62s | 1.16132 | 0.842166 | 0.0434028 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 18 | 46.80s | 1.15646 | 0.840656 | 0.0434028 | "
],
"text/plain": [
"| 18 | 46.80s | 1.15646 | 0.840656 | 0.0434028 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 19 | 48.95s | 1.15327 | 0.838994 | 0.0434028 | "
],
"text/plain": [
"| 19 | 48.95s | 1.15327 | 0.838994 | 0.0434028 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 20 | 51.21s | 1.15068 | 0.837869 | 0.0434028 | "
],
"text/plain": [
"| 20 | 51.21s | 1.15068 | 0.837869 | 0.0434028 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 21 | 53.40s | 1.14713 | 0.836618 | 0.0434028 | "
],
"text/plain": [
"| 21 | 53.40s | 1.14713 | 0.836618 | 0.0434028 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 22 | 55.60s | 1.14393 | 0.835341 | 0.0434028 | "
],
"text/plain": [
"| 22 | 55.60s | 1.14393 | 0.835341 | 0.0434028 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 23 | 57.78s | 1.14105 | 0.834355 | 0.0434028 | "
],
"text/plain": [
"| 23 | 57.78s | 1.14105 | 0.834355 | 0.0434028 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 24 | 59.92s | 1.13905 | 0.833327 | 0.0434028 | "
],
"text/plain": [
"| 24 | 59.92s | 1.13905 | 0.833327 | 0.0434028 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"| 25 | 1m 2s | 1.13617 | 0.832308 | 0.0434028 | "
],
"text/plain": [
"| 25 | 1m 2s | 1.13617 | 0.832308 | 0.0434028 |"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"+---------+--------------+-------------------+-----------------------+-------------+ "
],
"text/plain": [
"+---------+--------------+-------------------+-----------------------+-------------+"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"Optimization Complete: Maximum number of passes through the data reached. "
],
"text/plain": [
"Optimization Complete: Maximum number of passes through the data reached."
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"Computing final objective value and training RMSE. "
],
"text/plain": [
"Computing final objective value and training RMSE."
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
" Final objective value: 1.16266 "
],
"text/plain": [
" Final objective value: 1.16266"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
" Final training RMSE: 0.822791 "
],
"text/plain": [
" Final training RMSE: 0.822791"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"m_user = graphlab.recommender.create(train, 'user_id', 'movie_id', 'rating', \n",
" user_data=user_sf)\n",
"m_item = graphlab.recommender.create(train, 'user_id', 'movie_id', 'rating', \n",
" item_data=item_sf)\n",
"m_both = graphlab.recommender.create(train, 'user_id', 'movie_id', 'rating', \n",
" user_data=user_sf, item_data=item_sf)"
]
},
{
"cell_type": "code",
"execution_count": 31,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
"Class : RankingFactorizationRecommender\n",
"\n",
"Schema\n",
"------\n",
"User ID : user_id\n",
"Item ID : movie_id\n",
"Target : rating\n",
"Additional observation features : 1\n",
"Number of user side features : 5\n",
"Number of item side features : 3\n",
"\n",
"Statistics\n",
"----------\n",
"Number of observations : 965508\n",
"Number of users : 6040\n",
"Number of items : 3883\n",
"\n",
"Training summary\n",
"----------------\n",
"Training time : 74.1749\n",
"\n",
"Model Parameters\n",
"----------------\n",
"Model class : RankingFactorizationRecommender\n",
"num_factors : 32\n",
"binary_target : 0\n",
"side_data_factorization : 1\n",
"solver : auto\n",
"nmf : 0\n",
"max_iterations : 25\n",
"\n",
"Regularization Settings\n",
"-----------------------\n",
"regularization : 0.0\n",
"regularization_type : normal\n",
"linear_regularization : 0.0\n",
"ranking_regularization : 0.25\n",
"unobserved_rating_value : -1.79769313486e+308\n",
"num_sampled_negative_examples : 4\n",
"ials_confidence_scaling_type : auto\n",
"ials_confidence_scaling_factor : 1\n",
"\n",
"Optimization Settings\n",
"---------------------\n",
"init_random_sigma : 0.01\n",
"sgd_convergence_interval : 4\n",
"sgd_convergence_threshold : 0.0\n",
"sgd_max_trial_iterations : 5\n",
"sgd_sampling_block_size : 131072\n",
"sgd_step_adjustment_interval : 4\n",
"sgd_step_size : 0.0\n",
"sgd_trial_sample_minimum_size : 10000\n",
"sgd_trial_sample_proportion : 0.125\n",
"step_size_decrease_rate : 0.75\n",
"additional_iterations_if_unhealthy: 5\n",
"adagrad_momentum_weighting : 0.9\n",
"num_tempering_iterations : 4\n",
"tempering_regularization_start_value: 0.0\n",
"track_exact_loss : 0"
]
},
"execution_count": 31,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"m_both"
]
},
{
"cell_type": "code",
"execution_count": 32,
"metadata": {
"collapsed": false
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"compare_models: using 200 users to estimate model performance\n",
"PROGRESS: Evaluate model M0\n",
"\n",
"Precision and recall summary statistics by cutoff\n",
"+--------+----------------+-----------------+\n",
"| cutoff | mean_precision | mean_recall |\n",
"+--------+----------------+-----------------+\n",
"| 1 | 0.305 | 0.0104911692762 |\n",
"| 2 | 0.3075 | 0.0218598413174 |\n",
"| 3 | 0.298333333333 | 0.0312188865716 |\n",
"| 4 | 0.28375 | 0.0416981002021 |\n",
"| 5 | 0.267 | 0.0490821420076 |\n",
"| 6 | 0.255 | 0.0566601620839 |\n",
"| 7 | 0.244285714286 | 0.0634873210823 |\n",
"| 8 | 0.2325 | 0.0671075588864 |\n",
"| 9 | 0.227222222222 | 0.0732309783629 |\n",
"| 10 | 0.22 | 0.0778109177835 |\n",
"+--------+----------------+-----------------+\n",
"[10 rows x 3 columns]\n",
"\n",
"('\\nOverall RMSE: ', 0.9976486246840836)\n",
"\n",
"Per User RMSE (best)\n",
"+---------+-------+----------------+\n",
"| user_id | count | rmse |\n",
"+---------+-------+----------------+\n",
"| 4259 | 4 | 0.482071967797 |\n",
"+---------+-------+----------------+\n",
"[1 rows x 3 columns]\n",
"\n",
"\n",
"Per User RMSE (worst)\n",
"+---------+-------+---------------+\n",
"| user_id | count | rmse |\n",
"+---------+-------+---------------+\n",
"| 3275 | 4 | 1.92700170805 |\n",
"+---------+-------+---------------+\n",
"[1 rows x 3 columns]\n",
"\n",
"\n",
"Per Item RMSE (best)\n",
"+----------+-------+------------------+\n",
"| movie_id | count | rmse |\n",
"+----------+-------+------------------+\n",
"| 163 | 1 | 0.00169078452152 |\n",
"+----------+-------+------------------+\n",
"[1 rows x 3 columns]\n",
"\n",
"\n",
"Per Item RMSE (worst)\n",
"+----------+-------+---------------+\n",
"| movie_id | count | rmse |\n",
"+----------+-------+---------------+\n",
"| 3196 | 1 | 4.01836458102 |\n",
"+----------+-------+---------------+\n",
"[1 rows x 3 columns]\n",
"\n",
"PROGRESS: Evaluate model M1\n",
"\n",
"Precision and recall summary statistics by cutoff\n",
"+--------+----------------+-----------------+\n",
"| cutoff | mean_precision | mean_recall |\n",
"+--------+----------------+-----------------+\n",
"| 1 | 0.375 | 0.0136968745991 |\n",
"| 2 | 0.3575 | 0.023755002765 |\n",
"| 3 | 0.335 | 0.0362605125879 |\n",
"| 4 | 0.3075 | 0.0423008826942 |\n",
"| 5 | 0.293 | 0.0497532350863 |\n",
"| 6 | 0.281666666667 | 0.0578575934232 |\n",
"| 7 | 0.282142857143 | 0.0671244350417 |\n",
"| 8 | 0.27125 | 0.0748468085545 |\n",
"| 9 | 0.262222222222 | 0.0813855604447 |\n",
"| 10 | 0.26 | 0.0896391608922 |\n",
"+--------+----------------+-----------------+\n",
"[10 rows x 3 columns]\n",
"\n",
"('\\nOverall RMSE: ', 1.0433027431346846)\n",
"\n",
"Per User RMSE (best)\n",
"+---------+-------+---------------+\n",
"| user_id | count | rmse |\n",
"+---------+-------+---------------+\n",
"| 4259 | 4 | 0.24496854802 |\n",
"+---------+-------+---------------+\n",
"[1 rows x 3 columns]\n",
"\n",
"\n",
"Per User RMSE (worst)\n",
"+---------+-------+---------------+\n",
"| user_id | count | rmse |\n",
"+---------+-------+---------------+\n",
"| 2912 | 7 | 2.13582589594 |\n",
"+---------+-------+---------------+\n",
"[1 rows x 3 columns]\n",
"\n",
"\n",
"Per Item RMSE (best)\n",
"+----------+-------+------------------+\n",
"| movie_id | count | rmse |\n",
"+----------+-------+------------------+\n",
"| 379 | 1 | 0.00331903448671 |\n",
"+----------+-------+------------------+\n",
"[1 rows x 3 columns]\n",
"\n",
"\n",
"Per Item RMSE (worst)\n",
"+----------+-------+---------------+\n",
"| movie_id | count | rmse |\n",
"+----------+-------+---------------+\n",
"| 3117 | 1 | 3.61265547345 |\n",
"+----------+-------+---------------+\n",
"[1 rows x 3 columns]\n",
"\n",
"PROGRESS: Evaluate model M2\n",
"\n",
"Precision and recall summary statistics by cutoff\n",
"+--------+----------------+-----------------+\n",
"| cutoff | mean_precision | mean_recall |\n",
"+--------+----------------+-----------------+\n",
"| 1 | 0.355 | 0.01355116837 |\n",
"| 2 | 0.345 | 0.0247731072461 |\n",
"| 3 | 0.328333333333 | 0.0346428767624 |\n",
"| 4 | 0.3 | 0.0417405134962 |\n",
"| 5 | 0.286 | 0.0496330466776 |\n",
"| 6 | 0.283333333333 | 0.0607078786391 |\n",
"| 7 | 0.273571428571 | 0.068217750301 |\n",
"| 8 | 0.265625 | 0.0745364058279 |\n",
"| 9 | 0.262777777778 | 0.0821126915135 |\n",
"| 10 | 0.255 | 0.0868626759608 |\n",
"+--------+----------------+-----------------+\n",
"[10 rows x 3 columns]\n",
"\n",
"('\\nOverall RMSE: ', 1.0121070979184972)\n",
"\n",
"Per User RMSE (best)\n",
"+---------+-------+----------------+\n",
"| user_id | count | rmse |\n",
"+---------+-------+----------------+\n",
"| 4259 | 4 | 0.280887940151 |\n",
"+---------+-------+----------------+\n",
"[1 rows x 3 columns]\n",
"\n",
"\n",
"Per User RMSE (worst)\n",
"+---------+-------+---------------+\n",
"| user_id | count | rmse |\n",
"+---------+-------+---------------+\n",
"| 2912 | 7 | 2.21734756181 |\n",
"+---------+-------+---------------+\n",
"[1 rows x 3 columns]\n",
"\n",
"\n",
"Per Item RMSE (best)\n",
"+----------+-------+-------------------+\n",
"| movie_id | count | rmse |\n",
"+----------+-------+-------------------+\n",
"| 1283 | 1 | 0.000539099143753 |\n",
"+----------+-------+-------------------+\n",
"[1 rows x 3 columns]\n",
"\n",
"\n",
"Per Item RMSE (worst)\n",
"+----------+-------+---------------+\n",
"| movie_id | count | rmse |\n",
"+----------+-------+---------------+\n",
"| 1806 | 1 | 3.39939804512 |\n",
"+----------+-------+---------------+\n",
"[1 rows x 3 columns]\n",
"\n",
"PROGRESS: Evaluate model M3\n",
"\n",
"Precision and recall summary statistics by cutoff\n",
"+--------+----------------+-----------------+\n",
"| cutoff | mean_precision | mean_recall |\n",
"+--------+----------------+-----------------+\n",
"| 1 | 0.4 | 0.0173755798355 |\n",
"| 2 | 0.3625 | 0.0265019144168 |\n",
"| 3 | 0.338333333333 | 0.0380391678111 |\n",
"| 4 | 0.3275 | 0.0479764485634 |\n",
"| 5 | 0.312 | 0.0576493349033 |\n",
"| 6 | 0.299166666667 | 0.0654828586353 |\n",
"| 7 | 0.295714285714 | 0.0760004175285 |\n",
"| 8 | 0.29 | 0.0841766126368 |\n",
"| 9 | 0.286666666667 | 0.093854767667 |\n",
"| 10 | 0.2755 | 0.0995420682121 |\n",
"+--------+----------------+-----------------+\n",
"[10 rows x 3 columns]\n",
"\n",
"('\\nOverall RMSE: ', 0.9936036664127302)\n",
"\n",
"Per User RMSE (best)\n",
"+---------+-------+----------------+\n",
"| user_id | count | rmse |\n",
"+---------+-------+----------------+\n",
"| 4259 | 4 | 0.397063536039 |\n",
"+---------+-------+----------------+\n",
"[1 rows x 3 columns]\n",
"\n",
"\n",
"Per User RMSE (worst)\n",
"+---------+-------+---------------+\n",
"| user_id | count | rmse |\n",
"+---------+-------+---------------+\n",
"| 2912 | 7 | 2.01561012851 |\n",
"+---------+-------+---------------+\n",
"[1 rows x 3 columns]\n",
"\n",
"\n",
"Per Item RMSE (best)\n",
"+----------+-------+------------------+\n",
"| movie_id | count | rmse |\n",
"+----------+-------+------------------+\n",
"| 849 | 1 | 0.00179710693082 |\n",
"+----------+-------+------------------+\n",
"[1 rows x 3 columns]\n",
"\n",
"\n",
"Per Item RMSE (worst)\n",
"+----------+-------+---------------+\n",
"| movie_id | count | rmse |\n",
"+----------+-------+---------------+\n",
"| 3806 | 1 | 3.81246875723 |\n",
"+----------+-------+---------------+\n",
"[1 rows x 3 columns]\n",
"\n"
]
}
],
"source": [
"results = graphlab.recommender.util.compare_models(test, [m, m_user, m_item, m_both], user_sample=0.2)"
]
},
{
"cell_type": "code",
"execution_count": 33,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
"[0.9976486246840836,\n",
" 1.0433027431346846,\n",
" 1.0121070979184972,\n",
" 0.9936036664127302]"
]
},
"execution_count": 33,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"[results[i]['rmse_overall'] for i in range(len(results))]"
]
},
{
"cell_type": "code",
"execution_count": 34,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"application/javascript": [
"$(\"head\").append($(\"\"], \"descriptives\": {\"rows\": 2057, \"columns\": 3}, \"view_component\": \"Summary\", \"view_file\": \"sframe\", \"view_params\": {\"y\": null, \"x\": null, \"columns\": [\"movie_id\", \"count\", \"rmse\"], \"view\": null}, \"view_components\": [\"Summary\", \"Table\", \"Bar Chart\", \"BoxWhisker Plot\", \"Line Chart\", \"Scatter Plot\", \"Heat Map\", \"Plots\"], \"type\": \"SFrame\", \"columns\": [{\"dtype\": \"int\", \"name\": \"movie_id\"}, {\"dtype\": \"int\", \"name\": \"count\"}, {\"dtype\": \"float\", \"name\": \"rmse\"}], \"column_identifiers\": [\"count\", \"movie_id\", \"rmse\"]}, \"columns\": [{\"dtype\": \"int\", \"name\": \"movie_id\"}, {\"dtype\": \"int\", \"name\": \"count\"}, {\"dtype\": \"float\", \"name\": \"rmse\"}]}, e);\n",
" });\n",
" })();\n",
" "
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"results[0]['rmse_by_item'].show()"
]
},
{
"cell_type": "code",
"execution_count": 36,
"metadata": {
"collapsed": false
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"compare_models: using 183 users to estimate model performance\n",
"PROGRESS: Evaluate model M0\n",
"\n",
"Precision and recall summary statistics by cutoff\n",
"+--------+----------------+-----------------+\n",
"| cutoff | mean_precision | mean_recall |\n",
"+--------+----------------+-----------------+\n",
"| 1 | 0.158469945355 | 0.0235039678346 |\n",
"| 2 | 0.155737704918 | 0.0367981705878 |\n",
"| 3 | 0.136612021858 | 0.0436527349676 |\n",
"| 4 | 0.137978142077 | 0.0577237401993 |\n",
"| 5 | 0.136612021858 | 0.0723710331936 |\n",
"| 6 | 0.125683060109 | 0.0842016591703 |\n",
"| 7 | 0.124902419984 | 0.101563633319 |\n",
"| 8 | 0.118852459016 | 0.109983711543 |\n",
"| 9 | 0.114754098361 | 0.11908895883 |\n",
"| 10 | 0.111475409836 | 0.128536481539 |\n",
"+--------+----------------+-----------------+\n",
"[10 rows x 3 columns]\n",
"\n",
"PROGRESS: Evaluate model M1\n",
"\n",
"Precision and recall summary statistics by cutoff\n",
"+--------+----------------+-----------------+\n",
"| cutoff | mean_precision | mean_recall |\n",
"+--------+----------------+-----------------+\n",
"| 1 | 0.245901639344 | 0.0538736282089 |\n",
"| 2 | 0.234972677596 | 0.0758295229523 |\n",
"| 3 | 0.23679417122 | 0.101520870674 |\n",
"| 4 | 0.225409836066 | 0.12569838792 |\n",
"| 5 | 0.209836065574 | 0.142998857028 |\n",
"| 6 | 0.192167577413 | 0.155246605969 |\n",
"| 7 | 0.183450429352 | 0.174214946795 |\n",
"| 8 | 0.176229508197 | 0.190227369104 |\n",
"| 9 | 0.165148755313 | 0.200596309693 |\n",
"| 10 | 0.158469945355 | 0.210908575339 |\n",
"+--------+----------------+-----------------+\n",
"[10 rows x 3 columns]\n",
"\n"
]
},
{
"data": {
"text/plain": [
"[{'precision_recall_by_user': Columns:\n",
" \tuser_id\tint\n",
" \tcutoff\tint\n",
" \tprecision\tfloat\n",
" \trecall\tfloat\n",
" \tcount\tint\n",
" \n",
" Rows: 3294\n",
" \n",
" Data:\n",
" +---------+--------+-----------+--------+-------+\n",
" | user_id | cutoff | precision | recall | count |\n",
" +---------+--------+-----------+--------+-------+\n",
" | 3 | 1 | 0.0 | 0.0 | 2 |\n",
" | 3 | 2 | 0.0 | 0.0 | 2 |\n",
" | 3 | 3 | 0.0 | 0.0 | 2 |\n",
" | 3 | 4 | 0.0 | 0.0 | 2 |\n",
" | 3 | 5 | 0.0 | 0.0 | 2 |\n",
" | 3 | 6 | 0.0 | 0.0 | 2 |\n",
" | 3 | 7 | 0.0 | 0.0 | 2 |\n",
" | 3 | 8 | 0.0 | 0.0 | 2 |\n",
" | 3 | 9 | 0.0 | 0.0 | 2 |\n",
" | 3 | 10 | 0.0 | 0.0 | 2 |\n",
" +---------+--------+-----------+--------+-------+\n",
" [3294 rows x 5 columns]\n",
" Note: Only the head of the SFrame is printed.\n",
" You can use print_rows(num_rows=m, num_columns=n) to print more rows and columns.,\n",
" 'precision_recall_overall': Columns:\n",
" \tcutoff\tint\n",
" \tprecision\tfloat\n",
" \trecall\tfloat\n",
" \n",
" Rows: 18\n",
" \n",
" Data:\n",
" +--------+----------------+-----------------+\n",
" | cutoff | precision | recall |\n",
" +--------+----------------+-----------------+\n",
" | 1 | 0.158469945355 | 0.0235039678346 |\n",
" | 2 | 0.155737704918 | 0.0367981705878 |\n",
" | 3 | 0.136612021858 | 0.0436527349676 |\n",
" | 4 | 0.137978142077 | 0.0577237401993 |\n",
" | 5 | 0.136612021858 | 0.0723710331936 |\n",
" | 6 | 0.125683060109 | 0.0842016591703 |\n",
" | 7 | 0.124902419984 | 0.101563633319 |\n",
" | 8 | 0.118852459016 | 0.109983711543 |\n",
" | 9 | 0.114754098361 | 0.11908895883 |\n",
" | 10 | 0.111475409836 | 0.128536481539 |\n",
" +--------+----------------+-----------------+\n",
" [18 rows x 3 columns]\n",
" Note: Only the head of the SFrame is printed.\n",
" You can use print_rows(num_rows=m, num_columns=n) to print more rows and columns.},\n",
" {'precision_recall_by_user': Columns:\n",
" \tuser_id\tint\n",
" \tcutoff\tint\n",
" \tprecision\tfloat\n",
" \trecall\tfloat\n",
" \tcount\tint\n",
" \n",
" Rows: 3294\n",
" \n",
" Data:\n",
" +---------+--------+-----------+--------+-------+\n",
" | user_id | cutoff | precision | recall | count |\n",
" +---------+--------+-----------+--------+-------+\n",
" | 3 | 1 | 0.0 | 0.0 | 2 |\n",
" | 3 | 2 | 0.0 | 0.0 | 2 |\n",
" | 3 | 3 | 0.0 | 0.0 | 2 |\n",
" | 3 | 4 | 0.0 | 0.0 | 2 |\n",
" | 3 | 5 | 0.0 | 0.0 | 2 |\n",
" | 3 | 6 | 0.0 | 0.0 | 2 |\n",
" | 3 | 7 | 0.0 | 0.0 | 2 |\n",
" | 3 | 8 | 0.0 | 0.0 | 2 |\n",
" | 3 | 9 | 0.0 | 0.0 | 2 |\n",
" | 3 | 10 | 0.0 | 0.0 | 2 |\n",
" +---------+--------+-----------+--------+-------+\n",
" [3294 rows x 5 columns]\n",
" Note: Only the head of the SFrame is printed.\n",
" You can use print_rows(num_rows=m, num_columns=n) to print more rows and columns.,\n",
" 'precision_recall_overall': Columns:\n",
" \tcutoff\tint\n",
" \tprecision\tfloat\n",
" \trecall\tfloat\n",
" \n",
" Rows: 18\n",
" \n",
" Data:\n",
" +--------+----------------+-----------------+\n",
" | cutoff | precision | recall |\n",
" +--------+----------------+-----------------+\n",
" | 1 | 0.245901639344 | 0.0538736282089 |\n",
" | 2 | 0.234972677596 | 0.0758295229523 |\n",
" | 3 | 0.23679417122 | 0.101520870674 |\n",
" | 4 | 0.225409836066 | 0.12569838792 |\n",
" | 5 | 0.209836065574 | 0.142998857028 |\n",
" | 6 | 0.192167577413 | 0.155246605969 |\n",
" | 7 | 0.183450429352 | 0.174214946795 |\n",
" | 8 | 0.176229508197 | 0.190227369104 |\n",
" | 9 | 0.165148755313 | 0.200596309693 |\n",
" | 10 | 0.158469945355 | 0.210908575339 |\n",
" +--------+----------------+-----------------+\n",
" [18 rows x 3 columns]\n",
" Note: Only the head of the SFrame is printed.\n",
" You can use print_rows(num_rows=m, num_columns=n) to print more rows and columns.}]"
]
},
"execution_count": 36,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"graphlab.recommender.util.compare_models(test[test['rating'] > 4], \n",
" [m_rank, m_both], \n",
" user_sample=0.2, \n",
" metric='precision_recall')"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Factorization machines"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"collapsed": false
},
"outputs": [],
"source": [
"fm = graphlab.recommender.create(train.head(10000), 'user_id', 'movie_id', 'rating',\n",
" method='factorization_model',\n",
" item_data=item_sf, \n",
" sgd_step_size=0.09,\n",
" max_iterations=10)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"collapsed": false
},
"outputs": [],
"source": []
}
],
"metadata": {
"celltoolbar": "Slideshow",
"kernelspec": {
"display_name": "Python 2",
"language": "python",
"name": "python2"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 2
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython2",
"version": "2.7.11"
}
},
"nbformat": 4,
"nbformat_minor": 0
}