{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Optimization Methods\n",
"\n",
"Until now, you've always used Gradient Descent to update the parameters and minimize the cost. In this notebook, you will learn more advanced optimization methods that can speed up learning and perhaps even get you to a better final value for the cost function. Having a good optimization algorithm can be the difference between waiting days vs. just a few hours to get a good result. \n",
"\n",
"Gradient descent goes \"downhill\" on a cost function $J$. Think of it as trying to do this: \n",
"<img src=\"images/cost.jpg\" style=\"width:650px;height:300px;\">\n",
"<caption><center> <u> **Figure 1** </u>: **Minimizing the cost is like finding the lowest point in a hilly landscape**<br> At each step of the training, you update your parameters following a certain direction to try to get to the lowest possible point. </center></caption>\n",
"\n",
"**Notations**: As usual, $\\frac{\\partial J}{\\partial a } = $ `da` for any variable `a`.\n",
"\n",
"To get started, run the following code to import the libraries you will need."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### <font color='darkblue'> Updates to Assignment <font>\n",
"\n",
"#### If you were working on a previous version\n",
"* The current notebook filename is version \"Optimization_methods_v1b\". \n",
"* You can find your work in the file directory as version \"Optimization methods'.\n",
"* To see the file directory, click on the Coursera logo at the top left of the notebook.\n",
"\n",
"#### List of Updates\n",
"* op_utils is now opt_utils_v1a. Assertion statement in `initialize_parameters` is fixed.\n",
"* opt_utils_v1a: `compute_cost` function now accumulates total cost of the batch without taking the average (average is taken for entire epoch instead).\n",
"* In `model` function, the total cost per mini-batch is accumulated, and the average of the entire epoch is taken as the average cost. So the plot of the cost function over time is now a smooth downward curve instead of an oscillating curve.\n",
"* Print statements used to check each function are reformatted, and 'expected output` is reformatted to match the format of the print statements (for easier visual comparisons)."
]
},
{
"cell_type": "code",
"execution_count": 1,
"metadata": {
"collapsed": true
},
"outputs": [],
"source": [
"import numpy as np\n",
"import matplotlib.pyplot as plt\n",
"import scipy.io\n",
"import math\n",
"import sklearn\n",
"import sklearn.datasets\n",
"\n",
"from opt_utils_v1a import load_params_and_grads, initialize_parameters, forward_propagation, backward_propagation\n",
"from opt_utils_v1a import compute_cost, predict, predict_dec, plot_decision_boundary, load_dataset\n",
"from testCases import *\n",
"\n",
"%matplotlib inline\n",
"plt.rcParams['figure.figsize'] = (7.0, 4.0) # set default size of plots\n",
"plt.rcParams['image.interpolation'] = 'nearest'\n",
"plt.rcParams['image.cmap'] = 'gray'"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## 1 - Gradient Descent\n",
"\n",
"A simple optimization method in machine learning is gradient descent (GD). When you take gradient steps with respect to all $m$ examples on each step, it is also called Batch Gradient Descent. \n",
"\n",
"**Warm-up exercise**: Implement the gradient descent update rule. The gradient descent rule is, for $l = 1, ..., L$: \n",
"$$ W^{[l]} = W^{[l]} - \\alpha \\text{ } dW^{[l]} \\tag{1}$$\n",
"$$ b^{[l]} = b^{[l]} - \\alpha \\text{ } db^{[l]} \\tag{2}$$\n",
"\n",
"where L is the number of layers and $\\alpha$ is the learning rate. All parameters should be stored in the `parameters` dictionary. Note that the iterator `l` starts at 0 in the `for` loop while the first parameters are $W^{[1]}$ and $b^{[1]}$. You need to shift `l` to `l+1` when coding."
]
},
{
"cell_type": "code",
"execution_count": 2,
"metadata": {
"collapsed": true
},
"outputs": [],
"source": [
"# GRADED FUNCTION: update_parameters_with_gd\n",
"\n",
"def update_parameters_with_gd(parameters, grads, learning_rate):\n",
" \"\"\"\n",
" Update parameters using one step of gradient descent\n",
" \n",
" Arguments:\n",
" parameters -- python dictionary containing your parameters to be updated:\n",
" parameters['W' + str(l)] = Wl\n",
" parameters['b' + str(l)] = bl\n",
" grads -- python dictionary containing your gradients to update each parameters:\n",
" grads['dW' + str(l)] = dWl\n",
" grads['db' + str(l)] = dbl\n",
" learning_rate -- the learning rate, scalar.\n",
" \n",
" Returns:\n",
" parameters -- python dictionary containing your updated parameters \n",
" \"\"\"\n",
"\n",
" L = len(parameters) // 2 # number of layers in the neural networks\n",
"\n",
" # Update rule for each parameter\n",
" for l in range(L):\n",
" ### START CODE HERE ### (approx. 2 lines)\n",
" parameters[\"W\" + str(l+1)] = parameters[\"W\" + str(l + 1)] - learning_rate * grads[\"dW\" + str(l + 1)]\n",
" parameters[\"b\" + str(l+1)] = parameters[\"b\" + str(l + 1)] - learning_rate * grads[\"db\" + str(l + 1)]\n",
" ### END CODE HERE ###\n",
" \n",
" return parameters"
]
},
{
"cell_type": "code",
"execution_count": 3,
"metadata": {
"scrolled": true
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"W1 =\n",
"[[ 1.63535156 -0.62320365 -0.53718766]\n",
" [-1.07799357 0.85639907 -2.29470142]]\n",
"b1 =\n",
"[[ 1.74604067]\n",
" [-0.75184921]]\n",
"W2 =\n",
"[[ 0.32171798 -0.25467393 1.46902454]\n",
" [-2.05617317 -0.31554548 -0.3756023 ]\n",
" [ 1.1404819 -1.09976462 -0.1612551 ]]\n",
"b2 =\n",
"[[-0.88020257]\n",
" [ 0.02561572]\n",
" [ 0.57539477]]\n"
]
}
],
"source": [
"parameters, grads, learning_rate = update_parameters_with_gd_test_case()\n",
"\n",
"parameters = update_parameters_with_gd(parameters, grads, learning_rate)\n",
"print(\"W1 =\\n\" + str(parameters[\"W1\"]))\n",
"print(\"b1 =\\n\" + str(parameters[\"b1\"]))\n",
"print(\"W2 =\\n\" + str(parameters[\"W2\"]))\n",
"print(\"b2 =\\n\" + str(parameters[\"b2\"]))"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"**Expected Output**:\n",
"\n",
"```\n",
"W1 =\n",
"[[ 1.63535156 -0.62320365 -0.53718766]\n",
" [-1.07799357 0.85639907 -2.29470142]]\n",
"b1 =\n",
"[[ 1.74604067]\n",
" [-0.75184921]]\n",
"W2 =\n",
"[[ 0.32171798 -0.25467393 1.46902454]\n",
" [-2.05617317 -0.31554548 -0.3756023 ]\n",
" [ 1.1404819 -1.09976462 -0.1612551 ]]\n",
"b2 =\n",
"[[-0.88020257]\n",
" [ 0.02561572]\n",
" [ 0.57539477]]\n",
"```"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"A variant of this is Stochastic Gradient Descent (SGD), which is equivalent to mini-batch gradient descent where each mini-batch has just 1 example. The update rule that you have just implemented does not change. What changes is that you would be computing gradients on just one training example at a time, rather than on the whole training set. The code examples below illustrate the difference between stochastic gradient descent and (batch) gradient descent. \n",
"\n",
"- **(Batch) Gradient Descent**:\n",
"\n",
"``` python\n",
"X = data_input\n",
"Y = labels\n",
"parameters = initialize_parameters(layers_dims)\n",
"for i in range(0, num_iterations):\n",
" # Forward propagation\n",
" a, caches = forward_propagation(X, parameters)\n",
" # Compute cost.\n",
" cost += compute_cost(a, Y)\n",
" # Backward propagation.\n",
" grads = backward_propagation(a, caches, parameters)\n",
" # Update parameters.\n",
" parameters = update_parameters(parameters, grads)\n",
" \n",
"```\n",
"\n",
"- **Stochastic Gradient Descent**:\n",
"\n",
"```python\n",
"X = data_input\n",
"Y = labels\n",
"parameters = initialize_parameters(layers_dims)\n",
"for i in range(0, num_iterations):\n",
" for j in range(0, m):\n",
" # Forward propagation\n",
" a, caches = forward_propagation(X[:,j], parameters)\n",
" # Compute cost\n",
" cost += compute_cost(a, Y[:,j])\n",
" # Backward propagation\n",
" grads = backward_propagation(a, caches, parameters)\n",
" # Update parameters.\n",
" parameters = update_parameters(parameters, grads)\n",
"```\n"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"In Stochastic Gradient Descent, you use only 1 training example before updating the gradients. When the training set is large, SGD can be faster. But the parameters will \"oscillate\" toward the minimum rather than converge smoothly. Here is an illustration of this: \n",
"\n",
"<img src=\"images/kiank_sgd.png\" style=\"width:750px;height:250px;\">\n",
"<caption><center> <u> <font color='purple'> **Figure 1** </u><font color='purple'> : **SGD vs GD**<br> \"+\" denotes a minimum of the cost. SGD leads to many oscillations to reach convergence. But each step is a lot faster to compute for SGD than for GD, as it uses only one training example (vs. the whole batch for GD). </center></caption>\n",
"\n",
"**Note** also that implementing SGD requires 3 for-loops in total:\n",
"1. Over the number of iterations\n",
"2. Over the $m$ training examples\n",
"3. Over the layers (to update all parameters, from $(W^{[1]},b^{[1]})$ to $(W^{[L]},b^{[L]})$)\n",
"\n",
"In practice, you'll often get faster results if you do not use neither the whole training set, nor only one training example, to perform each update. Mini-batch gradient descent uses an intermediate number of examples for each step. With mini-batch gradient descent, you loop over the mini-batches instead of looping over individual training examples.\n",
"\n",
"<img src=\"images/kiank_minibatch.png\" style=\"width:750px;height:250px;\">\n",
"<caption><center> <u> <font color='purple'> **Figure 2** </u>: <font color='purple'> **SGD vs Mini-Batch GD**<br> \"+\" denotes a minimum of the cost. Using mini-batches in your optimization algorithm often leads to faster optimization. </center></caption>\n",
"\n",
"<font color='blue'>\n",
"**What you should remember**:\n",
"- The difference between gradient descent, mini-batch gradient descent and stochastic gradient descent is the number of examples you use to perform one update step.\n",
"- You have to tune a learning rate hyperparameter $\\alpha$.\n",
"- With a well-turned mini-batch size, usually it outperforms either gradient descent or stochastic gradient descent (particularly when the training set is large)."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## 2 - Mini-Batch Gradient descent\n",
"\n",
"Let's learn how to build mini-batches from the training set (X, Y).\n",
"\n",
"There are two steps:\n",
"- **Shuffle**: Create a shuffled version of the training set (X, Y) as shown below. Each column of X and Y represents a training example. Note that the random shuffling is done synchronously between X and Y. Such that after the shuffling the $i^{th}$ column of X is the example corresponding to the $i^{th}$ label in Y. The shuffling step ensures that examples will be split randomly into different mini-batches. \n",
"\n",
"<img src=\"images/kiank_shuffle.png\" style=\"width:550px;height:300px;\">\n",
"\n",
"- **Partition**: Partition the shuffled (X, Y) into mini-batches of size `mini_batch_size` (here 64). Note that the number of training examples is not always divisible by `mini_batch_size`. The last mini batch might be smaller, but you don't need to worry about this. When the final mini-batch is smaller than the full `mini_batch_size`, it will look like this: \n",
"\n",
"<img src=\"images/kiank_partition.png\" style=\"width:550px;height:300px;\">\n",
"\n",
"**Exercise**: Implement `random_mini_batches`. We coded the shuffling part for you. To help you with the partitioning step, we give you the following code that selects the indexes for the $1^{st}$ and $2^{nd}$ mini-batches:\n",
"```python\n",
"first_mini_batch_X = shuffled_X[:, 0 : mini_batch_size]\n",
"second_mini_batch_X = shuffled_X[:, mini_batch_size : 2 * mini_batch_size]\n",
"...\n",
"```\n",
"\n",
"Note that the last mini-batch might end up smaller than `mini_batch_size=64`. Let $\\lfloor s \\rfloor$ represents $s$ rounded down to the nearest integer (this is `math.floor(s)` in Python). If the total number of examples is not a multiple of `mini_batch_size=64` then there will be $\\lfloor \\frac{m}{mini\\_batch\\_size}\\rfloor$ mini-batches with a full 64 examples, and the number of examples in the final mini-batch will be ($m-mini_\\_batch_\\_size \\times \\lfloor \\frac{m}{mini\\_batch\\_size}\\rfloor$). "
]
},
{
"cell_type": "code",
"execution_count": 4,
"metadata": {
"collapsed": true
},
"outputs": [],
"source": [
"# GRADED FUNCTION: random_mini_batches\n",
"\n",
"def random_mini_batches(X, Y, mini_batch_size = 64, seed = 0):\n",
" \"\"\"\n",
" Creates a list of random minibatches from (X, Y)\n",
" \n",
" Arguments:\n",
" X -- input data, of shape (input size, number of examples)\n",
" Y -- true \"label\" vector (1 for blue dot / 0 for red dot), of shape (1, number of examples)\n",
" mini_batch_size -- size of the mini-batches, integer\n",
" \n",
" Returns:\n",
" mini_batches -- list of synchronous (mini_batch_X, mini_batch_Y)\n",
" \"\"\"\n",
" \n",
" np.random.seed(seed) # To make your \"random\" minibatches the same as ours\n",
" m = X.shape[1] # number of training examples\n",
" mini_batches = []\n",
" \n",
" # Step 1: Shuffle (X, Y)\n",
" permutation = list(np.random.permutation(m))\n",
" shuffled_X = X[:, permutation]\n",
" shuffled_Y = Y[:, permutation].reshape((1,m))\n",
"\n",
" # Step 2: Partition (shuffled_X, shuffled_Y). Minus the end case.\n",
" num_complete_minibatches = math.floor(m/mini_batch_size) # number of mini batches of size mini_batch_size in your partitionning\n",
" for k in range(0, num_complete_minibatches):\n",
" ### START CODE HERE ### (approx. 2 lines)\n",
" mini_batch_X = shuffled_X[:, k * mini_batch_size: (k + 1) * mini_batch_size]\n",
" mini_batch_Y = shuffled_Y[:, k * mini_batch_size: (k + 1) * mini_batch_size]\n",
" ### END CODE HERE ###\n",
" mini_batch = (mini_batch_X, mini_batch_Y)\n",
" mini_batches.append(mini_batch)\n",
" \n",
" # Handling the end case (last mini-batch < mini_batch_size)\n",
" if m % mini_batch_size != 0:\n",
" ### START CODE HERE ### (approx. 2 lines)\n",
" mini_batch_X = shuffled_X[:, -(m % mini_batch_size):]\n",
" mini_batch_Y = shuffled_Y[:, -(m % mini_batch_size):]\n",
" ### END CODE HERE ###\n",
" mini_batch = (mini_batch_X, mini_batch_Y)\n",
" mini_batches.append(mini_batch)\n",
" \n",
" return mini_batches"
]
},
{
"cell_type": "code",
"execution_count": 5,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"shape of the 1st mini_batch_X: (12288, 64)\n",
"shape of the 2nd mini_batch_X: (12288, 64)\n",
"shape of the 3rd mini_batch_X: (12288, 20)\n",
"shape of the 1st mini_batch_Y: (1, 64)\n",
"shape of the 2nd mini_batch_Y: (1, 64)\n",
"shape of the 3rd mini_batch_Y: (1, 20)\n",
"mini batch sanity check: [ 0.90085595 -0.7612069 0.2344157 ]\n"
]
}
],
"source": [
"X_assess, Y_assess, mini_batch_size = random_mini_batches_test_case()\n",
"mini_batches = random_mini_batches(X_assess, Y_assess, mini_batch_size)\n",
"\n",
"print (\"shape of the 1st mini_batch_X: \" + str(mini_batches[0][0].shape))\n",
"print (\"shape of the 2nd mini_batch_X: \" + str(mini_batches[1][0].shape))\n",
"print (\"shape of the 3rd mini_batch_X: \" + str(mini_batches[2][0].shape))\n",
"print (\"shape of the 1st mini_batch_Y: \" + str(mini_batches[0][1].shape))\n",
"print (\"shape of the 2nd mini_batch_Y: \" + str(mini_batches[1][1].shape)) \n",
"print (\"shape of the 3rd mini_batch_Y: \" + str(mini_batches[2][1].shape))\n",
"print (\"mini batch sanity check: \" + str(mini_batches[0][0][0][0:3]))"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"**Expected Output**:\n",
"\n",
"<table style=\"width:50%\"> \n",
" <tr>\n",
" <td > **shape of the 1st mini_batch_X** </td> \n",
" <td > (12288, 64) </td> \n",
" </tr> \n",
" \n",
" <tr>\n",
" <td > **shape of the 2nd mini_batch_X** </td> \n",
" <td > (12288, 64) </td> \n",
" </tr> \n",
" \n",
" <tr>\n",
" <td > **shape of the 3rd mini_batch_X** </td> \n",
" <td > (12288, 20) </td> \n",
" </tr>\n",
" <tr>\n",
" <td > **shape of the 1st mini_batch_Y** </td> \n",
" <td > (1, 64) </td> \n",
" </tr> \n",
" <tr>\n",
" <td > **shape of the 2nd mini_batch_Y** </td> \n",
" <td > (1, 64) </td> \n",
" </tr> \n",
" <tr>\n",
" <td > **shape of the 3rd mini_batch_Y** </td> \n",
" <td > (1, 20) </td> \n",
" </tr> \n",
" <tr>\n",
" <td > **mini batch sanity check** </td> \n",
" <td > [ 0.90085595 -0.7612069 0.2344157 ] </td> \n",
" </tr>\n",
" \n",
"</table>"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"<font color='blue'>\n",
"**What you should remember**:\n",
"- Shuffling and Partitioning are the two steps required to build mini-batches\n",
"- Powers of two are often chosen to be the mini-batch size, e.g., 16, 32, 64, 128."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## 3 - Momentum\n",
"\n",
"Because mini-batch gradient descent makes a parameter update after seeing just a subset of examples, the direction of the update has some variance, and so the path taken by mini-batch gradient descent will \"oscillate\" toward convergence. Using momentum can reduce these oscillations. \n",
"\n",
"Momentum takes into account the past gradients to smooth out the update. We will store the 'direction' of the previous gradients in the variable $v$. Formally, this will be the exponentially weighted average of the gradient on previous steps. You can also think of $v$ as the \"velocity\" of a ball rolling downhill, building up speed (and momentum) according to the direction of the gradient/slope of the hill. \n",
"\n",
"<img src=\"images/opt_momentum.png\" style=\"width:400px;height:250px;\">\n",
"<caption><center> <u><font color='purple'>**Figure 3**</u><font color='purple'>: The red arrows shows the direction taken by one step of mini-batch gradient descent with momentum. The blue points show the direction of the gradient (with respect to the current mini-batch) on each step. Rather than just following the gradient, we let the gradient influence $v$ and then take a step in the direction of $v$.<br> <font color='black'> </center>\n",
"\n",
"\n",
"**Exercise**: Initialize the velocity. The velocity, $v$, is a python dictionary that needs to be initialized with arrays of zeros. Its keys are the same as those in the `grads` dictionary, that is:\n",
"for $l =1,...,L$:\n",
"```python\n",
"v[\"dW\" + str(l+1)] = ... #(numpy array of zeros with the same shape as parameters[\"W\" + str(l+1)])\n",
"v[\"db\" + str(l+1)] = ... #(numpy array of zeros with the same shape as parameters[\"b\" + str(l+1)])\n",
"```\n",
"**Note** that the iterator l starts at 0 in the for loop while the first parameters are v[\"dW1\"] and v[\"db1\"] (that's a \"one\" on the superscript). This is why we are shifting l to l+1 in the `for` loop."
]
},
{
"cell_type": "code",
"execution_count": 6,
"metadata": {
"collapsed": true
},
"outputs": [],
"source": [
"# GRADED FUNCTION: initialize_velocity\n",
"\n",
"def initialize_velocity(parameters):\n",
" \"\"\"\n",
" Initializes the velocity as a python dictionary with:\n",
" - keys: \"dW1\", \"db1\", ..., \"dWL\", \"dbL\" \n",
" - values: numpy arrays of zeros of the same shape as the corresponding gradients/parameters.\n",
" Arguments:\n",
" parameters -- python dictionary containing your parameters.\n",
" parameters['W' + str(l)] = Wl\n",
" parameters['b' + str(l)] = bl\n",
" \n",
" Returns:\n",
" v -- python dictionary containing the current velocity.\n",
" v['dW' + str(l)] = velocity of dWl\n",
" v['db' + str(l)] = velocity of dbl\n",
" \"\"\"\n",
" \n",
" L = len(parameters) // 2 # number of layers in the neural networks\n",
" v = {}\n",
" \n",
" # Initialize velocity\n",
" for l in range(L):\n",
" ### START CODE HERE ### (approx. 2 lines)\n",
" v[\"dW\" + str(l+1)] = np.zeros_like(parameters['W' + str(l + 1)])\n",
" v[\"db\" + str(l+1)] = np.zeros_like(parameters['b' + str(l + 1)])\n",
" ### END CODE HERE ###\n",
" \n",
" return v"
]
},
{
"cell_type": "code",
"execution_count": 7,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"v[\"dW1\"] =\n",
"[[ 0. 0. 0.]\n",
" [ 0. 0. 0.]]\n",
"v[\"db1\"] =\n",
"[[ 0.]\n",
" [ 0.]]\n",
"v[\"dW2\"] =\n",
"[[ 0. 0. 0.]\n",
" [ 0. 0. 0.]\n",
" [ 0. 0. 0.]]\n",
"v[\"db2\"] =\n",
"[[ 0.]\n",
" [ 0.]\n",
" [ 0.]]\n"
]
}
],
"source": [
"parameters = initialize_velocity_test_case()\n",
"\n",
"v = initialize_velocity(parameters)\n",
"print(\"v[\\\"dW1\\\"] =\\n\" + str(v[\"dW1\"]))\n",
"print(\"v[\\\"db1\\\"] =\\n\" + str(v[\"db1\"]))\n",
"print(\"v[\\\"dW2\\\"] =\\n\" + str(v[\"dW2\"]))\n",
"print(\"v[\\\"db2\\\"] =\\n\" + str(v[\"db2\"]))"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"**Expected Output**:\n",
"\n",
"```\n",
"v[\"dW1\"] =\n",
"[[ 0. 0. 0.]\n",
" [ 0. 0. 0.]]\n",
"v[\"db1\"] =\n",
"[[ 0.]\n",
" [ 0.]]\n",
"v[\"dW2\"] =\n",
"[[ 0. 0. 0.]\n",
" [ 0. 0. 0.]\n",
" [ 0. 0. 0.]]\n",
"v[\"db2\"] =\n",
"[[ 0.]\n",
" [ 0.]\n",
" [ 0.]]\n",
"```"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"**Exercise**: Now, implement the parameters update with momentum. The momentum update rule is, for $l = 1, ..., L$: \n",
"\n",
"$$ \\begin{cases}\n",
"v_{dW^{[l]}} = \\beta v_{dW^{[l]}} + (1 - \\beta) dW^{[l]} \\\\\n",
"W^{[l]} = W^{[l]} - \\alpha v_{dW^{[l]}}\n",
"\\end{cases}\\tag{3}$$\n",
"\n",
"$$\\begin{cases}\n",
"v_{db^{[l]}} = \\beta v_{db^{[l]}} + (1 - \\beta) db^{[l]} \\\\\n",
"b^{[l]} = b^{[l]} - \\alpha v_{db^{[l]}} \n",
"\\end{cases}\\tag{4}$$\n",
"\n",
"where L is the number of layers, $\\beta$ is the momentum and $\\alpha$ is the learning rate. All parameters should be stored in the `parameters` dictionary. Note that the iterator `l` starts at 0 in the `for` loop while the first parameters are $W^{[1]}$ and $b^{[1]}$ (that's a \"one\" on the superscript). So you will need to shift `l` to `l+1` when coding."
]
},
{
"cell_type": "code",
"execution_count": 8,
"metadata": {
"collapsed": true
},
"outputs": [],
"source": [
"# GRADED FUNCTION: update_parameters_with_momentum\n",
"\n",
"def update_parameters_with_momentum(parameters, grads, v, beta, learning_rate):\n",
" \"\"\"\n",
" Update parameters using Momentum\n",
" \n",
" Arguments:\n",
" parameters -- python dictionary containing your parameters:\n",
" parameters['W' + str(l)] = Wl\n",
" parameters['b' + str(l)] = bl\n",
" grads -- python dictionary containing your gradients for each parameters:\n",
" grads['dW' + str(l)] = dWl\n",
" grads['db' + str(l)] = dbl\n",
" v -- python dictionary containing the current velocity:\n",
" v['dW' + str(l)] = ...\n",
" v['db' + str(l)] = ...\n",
" beta -- the momentum hyperparameter, scalar\n",
" learning_rate -- the learning rate, scalar\n",
" \n",
" Returns:\n",
" parameters -- python dictionary containing your updated parameters \n",
" v -- python dictionary containing your updated velocities\n",
" \"\"\"\n",
"\n",
" L = len(parameters) // 2 # number of layers in the neural networks\n",
" \n",
" # Momentum update for each parameter\n",
" for l in range(L):\n",
" \n",
" ### START CODE HERE ### (approx. 4 lines)\n",
" # compute velocities\n",
" v[\"dW\" + str(l+1)] = beta * v[\"dW\" + str(l + 1)] + (1 - beta) * grads['dW' + str(l + 1)]\n",
" v[\"db\" + str(l+1)] = beta * v[\"db\" + str(l + 1)] + (1 - beta) * grads['db' + str(l + 1)]\n",
" # update parameters\n",
" parameters[\"W\" + str(l+1)] = parameters[\"W\" + str(l + 1)] - learning_rate * v[\"dW\" + str(l + 1)]\n",
" parameters[\"b\" + str(l+1)] = parameters[\"b\" + str(l + 1)] - learning_rate * v[\"db\" + str(l + 1)]\n",
" ### END CODE HERE ###\n",
" \n",
" return parameters, v"
]
},
{
"cell_type": "code",
"execution_count": 9,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"W1 = \n",
"[[ 1.62544598 -0.61290114 -0.52907334]\n",
" [-1.07347112 0.86450677 -2.30085497]]\n",
"b1 = \n",
"[[ 1.74493465]\n",
" [-0.76027113]]\n",
"W2 = \n",
"[[ 0.31930698 -0.24990073 1.4627996 ]\n",
" [-2.05974396 -0.32173003 -0.38320915]\n",
" [ 1.13444069 -1.0998786 -0.1713109 ]]\n",
"b2 = \n",
"[[-0.87809283]\n",
" [ 0.04055394]\n",
" [ 0.58207317]]\n",
"v[\"dW1\"] = \n",
"[[-0.11006192 0.11447237 0.09015907]\n",
" [ 0.05024943 0.09008559 -0.06837279]]\n",
"v[\"db1\"] = \n",
"[[-0.01228902]\n",
" [-0.09357694]]\n",
"v[\"dW2\"] = \n",
"[[-0.02678881 0.05303555 -0.06916608]\n",
" [-0.03967535 -0.06871727 -0.08452056]\n",
" [-0.06712461 -0.00126646 -0.11173103]]\n",
"v[\"db2\"] = v[[ 0.02344157]\n",
" [ 0.16598022]\n",
" [ 0.07420442]]\n"
]
}
],
"source": [
"parameters, grads, v = update_parameters_with_momentum_test_case()\n",
"\n",
"parameters, v = update_parameters_with_momentum(parameters, grads, v, beta = 0.9, learning_rate = 0.01)\n",
"print(\"W1 = \\n\" + str(parameters[\"W1\"]))\n",
"print(\"b1 = \\n\" + str(parameters[\"b1\"]))\n",
"print(\"W2 = \\n\" + str(parameters[\"W2\"]))\n",
"print(\"b2 = \\n\" + str(parameters[\"b2\"]))\n",
"print(\"v[\\\"dW1\\\"] = \\n\" + str(v[\"dW1\"]))\n",
"print(\"v[\\\"db1\\\"] = \\n\" + str(v[\"db1\"]))\n",
"print(\"v[\\\"dW2\\\"] = \\n\" + str(v[\"dW2\"]))\n",
"print(\"v[\\\"db2\\\"] = v\" + str(v[\"db2\"]))"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"**Expected Output**:\n",
"\n",
"```\n",
"W1 = \n",
"[[ 1.62544598 -0.61290114 -0.52907334]\n",
" [-1.07347112 0.86450677 -2.30085497]]\n",
"b1 = \n",
"[[ 1.74493465]\n",
" [-0.76027113]]\n",
"W2 = \n",
"[[ 0.31930698 -0.24990073 1.4627996 ]\n",
" [-2.05974396 -0.32173003 -0.38320915]\n",
" [ 1.13444069 -1.0998786 -0.1713109 ]]\n",
"b2 = \n",
"[[-0.87809283]\n",
" [ 0.04055394]\n",
" [ 0.58207317]]\n",
"v[\"dW1\"] = \n",
"[[-0.11006192 0.11447237 0.09015907]\n",
" [ 0.05024943 0.09008559 -0.06837279]]\n",
"v[\"db1\"] = \n",
"[[-0.01228902]\n",
" [-0.09357694]]\n",
"v[\"dW2\"] = \n",
"[[-0.02678881 0.05303555 -0.06916608]\n",
" [-0.03967535 -0.06871727 -0.08452056]\n",
" [-0.06712461 -0.00126646 -0.11173103]]\n",
"v[\"db2\"] = v[[ 0.02344157]\n",
" [ 0.16598022]\n",
" [ 0.07420442]]\n",
"```"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"**Note** that:\n",
"- The velocity is initialized with zeros. So the algorithm will take a few iterations to \"build up\" velocity and start to take bigger steps.\n",
"- If $\\beta = 0$, then this just becomes standard gradient descent without momentum. \n",
"\n",
"**How do you choose $\\beta$?**\n",
"\n",
"- The larger the momentum $\\beta$ is, the smoother the update because the more we take the past gradients into account. But if $\\beta$ is too big, it could also smooth out the updates too much. \n",
"- Common values for $\\beta$ range from 0.8 to 0.999. If you don't feel inclined to tune this, $\\beta = 0.9$ is often a reasonable default. \n",
"- Tuning the optimal $\\beta$ for your model might need trying several values to see what works best in term of reducing the value of the cost function $J$. "
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"<font color='blue'>\n",
"**What you should remember**:\n",
"- Momentum takes past gradients into account to smooth out the steps of gradient descent. It can be applied with batch gradient descent, mini-batch gradient descent or stochastic gradient descent.\n",
"- You have to tune a momentum hyperparameter $\\beta$ and a learning rate $\\alpha$."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## 4 - Adam\n",
"\n",
"Adam is one of the most effective optimization algorithms for training neural networks. It combines ideas from RMSProp (described in lecture) and Momentum. \n",
"\n",
"**How does Adam work?**\n",
"1. It calculates an exponentially weighted average of past gradients, and stores it in variables $v$ (before bias correction) and $v^{corrected}$ (with bias correction). \n",
"2. It calculates an exponentially weighted average of the squares of the past gradients, and stores it in variables $s$ (before bias correction) and $s^{corrected}$ (with bias correction). \n",
"3. It updates parameters in a direction based on combining information from \"1\" and \"2\".\n",
"\n",
"The update rule is, for $l = 1, ..., L$: \n",
"\n",
"$$\\begin{cases}\n",
"v_{dW^{[l]}} = \\beta_1 v_{dW^{[l]}} + (1 - \\beta_1) \\frac{\\partial \\mathcal{J} }{ \\partial W^{[l]} } \\\\\n",
"v^{corrected}_{dW^{[l]}} = \\frac{v_{dW^{[l]}}}{1 - (\\beta_1)^t} \\\\\n",
"s_{dW^{[l]}} = \\beta_2 s_{dW^{[l]}} + (1 - \\beta_2) (\\frac{\\partial \\mathcal{J} }{\\partial W^{[l]} })^2 \\\\\n",
"s^{corrected}_{dW^{[l]}} = \\frac{s_{dW^{[l]}}}{1 - (\\beta_2)^t} \\\\\n",
"W^{[l]} = W^{[l]} - \\alpha \\frac{v^{corrected}_{dW^{[l]}}}{\\sqrt{s^{corrected}_{dW^{[l]}}} + \\varepsilon}\n",
"\\end{cases}$$\n",
"where:\n",
"- t counts the number of steps taken of Adam \n",
"- L is the number of layers\n",
"- $\\beta_1$ and $\\beta_2$ are hyperparameters that control the two exponentially weighted averages. \n",
"- $\\alpha$ is the learning rate\n",
"- $\\varepsilon$ is a very small number to avoid dividing by zero\n",
"\n",
"As usual, we will store all parameters in the `parameters` dictionary "
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"**Exercise**: Initialize the Adam variables $v, s$ which keep track of the past information.\n",
"\n",
"**Instruction**: The variables $v, s$ are python dictionaries that need to be initialized with arrays of zeros. Their keys are the same as for `grads`, that is:\n",
"for $l = 1, ..., L$:\n",
"```python\n",
"v[\"dW\" + str(l+1)] = ... #(numpy array of zeros with the same shape as parameters[\"W\" + str(l+1)])\n",
"v[\"db\" + str(l+1)] = ... #(numpy array of zeros with the same shape as parameters[\"b\" + str(l+1)])\n",
"s[\"dW\" + str(l+1)] = ... #(numpy array of zeros with the same shape as parameters[\"W\" + str(l+1)])\n",
"s[\"db\" + str(l+1)] = ... #(numpy array of zeros with the same shape as parameters[\"b\" + str(l+1)])\n",
"\n",
"```"
]
},
{
"cell_type": "code",
"execution_count": 10,
"metadata": {
"collapsed": true
},
"outputs": [],
"source": [
"# GRADED FUNCTION: initialize_adam\n",
"\n",
"def initialize_adam(parameters) :\n",
" \"\"\"\n",
" Initializes v and s as two python dictionaries with:\n",
" - keys: \"dW1\", \"db1\", ..., \"dWL\", \"dbL\" \n",
" - values: numpy arrays of zeros of the same shape as the corresponding gradients/parameters.\n",
" \n",
" Arguments:\n",
" parameters -- python dictionary containing your parameters.\n",
" parameters[\"W\" + str(l)] = Wl\n",
" parameters[\"b\" + str(l)] = bl\n",
" \n",
" Returns: \n",
" v -- python dictionary that will contain the exponentially weighted average of the gradient.\n",
" v[\"dW\" + str(l)] = ...\n",
" v[\"db\" + str(l)] = ...\n",
" s -- python dictionary that will contain the exponentially weighted average of the squared gradient.\n",
" s[\"dW\" + str(l)] = ...\n",
" s[\"db\" + str(l)] = ...\n",
"\n",
" \"\"\"\n",
" \n",
" L = len(parameters) // 2 # number of layers in the neural networks\n",
" v = {}\n",
" s = {}\n",
" \n",
" # Initialize v, s. Input: \"parameters\". Outputs: \"v, s\".\n",
" for l in range(L):\n",
" ### START CODE HERE ### (approx. 4 lines)\n",
" v[\"dW\" + str(l+1)] = np.zeros_like(parameters[\"W\" + str(l + 1)])\n",
" v[\"db\" + str(l+1)] = np.zeros_like(parameters[\"b\" + str(l + 1)])\n",
" s[\"dW\" + str(l+1)] = np.zeros_like(parameters[\"W\" + str(l + 1)])\n",
" s[\"db\" + str(l+1)] = np.zeros_like(parameters[\"b\" + str(l + 1)])\n",
" ### END CODE HERE ###\n",
" \n",
" return v, s"
]
},
{
"cell_type": "code",
"execution_count": 11,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"v[\"dW1\"] = \n",
"[[ 0. 0. 0.]\n",
" [ 0. 0. 0.]]\n",
"v[\"db1\"] = \n",
"[[ 0.]\n",
" [ 0.]]\n",
"v[\"dW2\"] = \n",
"[[ 0. 0. 0.]\n",
" [ 0. 0. 0.]\n",
" [ 0. 0. 0.]]\n",
"v[\"db2\"] = \n",
"[[ 0.]\n",
" [ 0.]\n",
" [ 0.]]\n",
"s[\"dW1\"] = \n",
"[[ 0. 0. 0.]\n",
" [ 0. 0. 0.]]\n",
"s[\"db1\"] = \n",
"[[ 0.]\n",
" [ 0.]]\n",
"s[\"dW2\"] = \n",
"[[ 0. 0. 0.]\n",
" [ 0. 0. 0.]\n",
" [ 0. 0. 0.]]\n",
"s[\"db2\"] = \n",
"[[ 0.]\n",
" [ 0.]\n",
" [ 0.]]\n"
]
}
],
"source": [
"parameters = initialize_adam_test_case()\n",
"\n",
"v, s = initialize_adam(parameters)\n",
"print(\"v[\\\"dW1\\\"] = \\n\" + str(v[\"dW1\"]))\n",
"print(\"v[\\\"db1\\\"] = \\n\" + str(v[\"db1\"]))\n",
"print(\"v[\\\"dW2\\\"] = \\n\" + str(v[\"dW2\"]))\n",
"print(\"v[\\\"db2\\\"] = \\n\" + str(v[\"db2\"]))\n",
"print(\"s[\\\"dW1\\\"] = \\n\" + str(s[\"dW1\"]))\n",
"print(\"s[\\\"db1\\\"] = \\n\" + str(s[\"db1\"]))\n",
"print(\"s[\\\"dW2\\\"] = \\n\" + str(s[\"dW2\"]))\n",
"print(\"s[\\\"db2\\\"] = \\n\" + str(s[\"db2\"]))"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"**Expected Output**:\n",
"\n",
"```\n",
"v[\"dW1\"] = \n",
"[[ 0. 0. 0.]\n",
" [ 0. 0. 0.]]\n",
"v[\"db1\"] = \n",
"[[ 0.]\n",
" [ 0.]]\n",
"v[\"dW2\"] = \n",
"[[ 0. 0. 0.]\n",
" [ 0. 0. 0.]\n",
" [ 0. 0. 0.]]\n",
"v[\"db2\"] = \n",
"[[ 0.]\n",
" [ 0.]\n",
" [ 0.]]\n",
"s[\"dW1\"] = \n",
"[[ 0. 0. 0.]\n",
" [ 0. 0. 0.]]\n",
"s[\"db1\"] = \n",
"[[ 0.]\n",
" [ 0.]]\n",
"s[\"dW2\"] = \n",
"[[ 0. 0. 0.]\n",
" [ 0. 0. 0.]\n",
" [ 0. 0. 0.]]\n",
"s[\"db2\"] = \n",
"[[ 0.]\n",
" [ 0.]\n",
" [ 0.]]\n",
"```"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"**Exercise**: Now, implement the parameters update with Adam. Recall the general update rule is, for $l = 1, ..., L$: \n",
"\n",
"$$\\begin{cases}\n",
"v_{W^{[l]}} = \\beta_1 v_{W^{[l]}} + (1 - \\beta_1) \\frac{\\partial J }{ \\partial W^{[l]} } \\\\\n",
"v^{corrected}_{W^{[l]}} = \\frac{v_{W^{[l]}}}{1 - (\\beta_1)^t} \\\\\n",
"s_{W^{[l]}} = \\beta_2 s_{W^{[l]}} + (1 - \\beta_2) (\\frac{\\partial J }{\\partial W^{[l]} })^2 \\\\\n",
"s^{corrected}_{W^{[l]}} = \\frac{s_{W^{[l]}}}{1 - (\\beta_2)^t} \\\\\n",
"W^{[l]} = W^{[l]} - \\alpha \\frac{v^{corrected}_{W^{[l]}}}{\\sqrt{s^{corrected}_{W^{[l]}}}+\\varepsilon}\n",
"\\end{cases}$$\n",
"\n",
"\n",
"**Note** that the iterator `l` starts at 0 in the `for` loop while the first parameters are $W^{[1]}$ and $b^{[1]}$. You need to shift `l` to `l+1` when coding."
]
},
{
"cell_type": "code",
"execution_count": 12,
"metadata": {
"collapsed": true
},
"outputs": [],
"source": [
"# GRADED FUNCTION: update_parameters_with_adam\n",
"\n",
"def update_parameters_with_adam(parameters, grads, v, s, t, learning_rate = 0.01,\n",
" beta1 = 0.9, beta2 = 0.999, epsilon = 1e-8):\n",
" \"\"\"\n",
" Update parameters using Adam\n",
" \n",
" Arguments:\n",
" parameters -- python dictionary containing your parameters:\n",
" parameters['W' + str(l)] = Wl\n",
" parameters['b' + str(l)] = bl\n",
" grads -- python dictionary containing your gradients for each parameters:\n",
" grads['dW' + str(l)] = dWl\n",
" grads['db' + str(l)] = dbl\n",
" v -- Adam variable, moving average of the first gradient, python dictionary\n",
" s -- Adam variable, moving average of the squared gradient, python dictionary\n",
" learning_rate -- the learning rate, scalar.\n",
" beta1 -- Exponential decay hyperparameter for the first moment estimates \n",
" beta2 -- Exponential decay hyperparameter for the second moment estimates \n",
" epsilon -- hyperparameter preventing division by zero in Adam updates\n",
"\n",
" Returns:\n",
" parameters -- python dictionary containing your updated parameters \n",
" v -- Adam variable, moving average of the first gradient, python dictionary\n",
" s -- Adam variable, moving average of the squared gradient, python dictionary\n",
" \"\"\"\n",
" \n",
" L = len(parameters) // 2 # number of layers in the neural networks\n",
" v_corrected = {} # Initializing first moment estimate, python dictionary\n",
" s_corrected = {} # Initializing second moment estimate, python dictionary\n",
" \n",
" # Perform Adam update on all parameters\n",
" for l in range(L):\n",
" # Moving average of the gradients. Inputs: \"v, grads, beta1\". Output: \"v\".\n",
" ### START CODE HERE ### (approx. 2 lines)\n",
" v[\"dW\" + str(l+1)] = beta1 * v[\"dW\" + str(l + 1)] + (1 - beta1) * grads[\"dW\" + str(l + 1)]\n",
" v[\"db\" + str(l+1)] = beta1 * v[\"db\" + str(l + 1)] + (1 - beta1) * grads[\"db\" + str(l + 1)]\n",
" ### END CODE HERE ###\n",
"\n",
" # Compute bias-corrected first moment estimate. Inputs: \"v, beta1, t\". Output: \"v_corrected\".\n",
" ### START CODE HERE ### (approx. 2 lines)\n",
" v_corrected[\"dW\" + str(l+1)] = v[\"dW\" + str(l + 1)] / (1 - beta1 ** t)\n",
" v_corrected[\"db\" + str(l+1)] = v[\"db\" + str(l + 1)] / (1 - beta1 ** t)\n",
" ### END CODE HERE ###\n",
"\n",
" # Moving average of the squared gradients. Inputs: \"s, grads, beta2\". Output: \"s\".\n",
" ### START CODE HERE ### (approx. 2 lines)\n",
" s[\"dW\" + str(l+1)] = beta2 * s[\"dW\" + str(l + 1)] + (1 - beta2) * (grads[\"dW\" + str(l + 1)] ** 2)\n",
" s[\"db\" + str(l+1)] = beta2 * s[\"db\" + str(l + 1)] + (1 - beta2) * (grads[\"db\" + str(l + 1)] ** 2)\n",
" ### END CODE HERE ###\n",
"\n",
" # Compute bias-corrected second raw moment estimate. Inputs: \"s, beta2, t\". Output: \"s_corrected\".\n",
" ### START CODE HERE ### (approx. 2 lines)\n",
" s_corrected[\"dW\" + str(l+1)] = s[\"dW\" + str(l + 1)] / (1 - beta2 ** t)\n",
" s_corrected[\"db\" + str(l+1)] = s[\"db\" + str(l + 1)] / (1 - beta2 ** t)\n",
" ### END CODE HERE ###\n",
"\n",
" # Update parameters. Inputs: \"parameters, learning_rate, v_corrected, s_corrected, epsilon\". Output: \"parameters\".\n",
" ### START CODE HERE ### (approx. 2 lines)\n",
" parameters[\"W\" + str(l+1)] = (parameters[\"W\" + str(l + 1)] - learning_rate *\n",
" v_corrected[\"dW\" + str(l + 1)] / (np.sqrt(s_corrected[\"dW\" + str(l + 1)]) + epsilon))\n",
" parameters[\"b\" + str(l+1)] = (parameters[\"b\" + str(l + 1)] - learning_rate *\n",
" v_corrected[\"db\" + str(l + 1)] / (np.sqrt(s_corrected[\"db\" + str(l + 1)]) + epsilon))\n",
" ### END CODE HERE ###\n",
"\n",
" return parameters, v, s"
]
},
{
"cell_type": "code",
"execution_count": 13,
"metadata": {
"scrolled": false
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"W1 = \n",
"[[ 1.63178673 -0.61919778 -0.53561312]\n",
" [-1.08040999 0.85796626 -2.29409733]]\n",
"b1 = \n",
"[[ 1.75225313]\n",
" [-0.75376553]]\n",
"W2 = \n",
"[[ 0.32648046 -0.25681174 1.46954931]\n",
" [-2.05269934 -0.31497584 -0.37661299]\n",
" [ 1.14121081 -1.09244991 -0.16498684]]\n",
"b2 = \n",
"[[-0.88529979]\n",
" [ 0.03477238]\n",
" [ 0.57537385]]\n",
"v[\"dW1\"] = \n",
"[[-0.11006192 0.11447237 0.09015907]\n",
" [ 0.05024943 0.09008559 -0.06837279]]\n",
"v[\"db1\"] = \n",
"[[-0.01228902]\n",
" [-0.09357694]]\n",
"v[\"dW2\"] = \n",
"[[-0.02678881 0.05303555 -0.06916608]\n",
" [-0.03967535 -0.06871727 -0.08452056]\n",
" [-0.06712461 -0.00126646 -0.11173103]]\n",
"v[\"db2\"] = \n",
"[[ 0.02344157]\n",
" [ 0.16598022]\n",
" [ 0.07420442]]\n",
"s[\"dW1\"] = \n",
"[[ 0.00121136 0.00131039 0.00081287]\n",
" [ 0.0002525 0.00081154 0.00046748]]\n",
"s[\"db1\"] = \n",
"[[ 1.51020075e-05]\n",
" [ 8.75664434e-04]]\n",
"s[\"dW2\"] = \n",
"[[ 7.17640232e-05 2.81276921e-04 4.78394595e-04]\n",
" [ 1.57413361e-04 4.72206320e-04 7.14372576e-04]\n",
" [ 4.50571368e-04 1.60392066e-07 1.24838242e-03]]\n",
"s[\"db2\"] = \n",
"[[ 5.49507194e-05]\n",
" [ 2.75494327e-03]\n",
" [ 5.50629536e-04]]\n"
]
}
],
"source": [
"parameters, grads, v, s = update_parameters_with_adam_test_case()\n",
"parameters, v, s = update_parameters_with_adam(parameters, grads, v, s, t = 2)\n",
"\n",
"print(\"W1 = \\n\" + str(parameters[\"W1\"]))\n",
"print(\"b1 = \\n\" + str(parameters[\"b1\"]))\n",
"print(\"W2 = \\n\" + str(parameters[\"W2\"]))\n",
"print(\"b2 = \\n\" + str(parameters[\"b2\"]))\n",
"print(\"v[\\\"dW1\\\"] = \\n\" + str(v[\"dW1\"]))\n",
"print(\"v[\\\"db1\\\"] = \\n\" + str(v[\"db1\"]))\n",
"print(\"v[\\\"dW2\\\"] = \\n\" + str(v[\"dW2\"]))\n",
"print(\"v[\\\"db2\\\"] = \\n\" + str(v[\"db2\"]))\n",
"print(\"s[\\\"dW1\\\"] = \\n\" + str(s[\"dW1\"]))\n",
"print(\"s[\\\"db1\\\"] = \\n\" + str(s[\"db1\"]))\n",
"print(\"s[\\\"dW2\\\"] = \\n\" + str(s[\"dW2\"]))\n",
"print(\"s[\\\"db2\\\"] = \\n\" + str(s[\"db2\"]))"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"**Expected Output**:\n",
"\n",
"```\n",
"W1 = \n",
"[[ 1.63178673 -0.61919778 -0.53561312]\n",
" [-1.08040999 0.85796626 -2.29409733]]\n",
"b1 = \n",
"[[ 1.75225313]\n",
" [-0.75376553]]\n",
"W2 = \n",
"[[ 0.32648046 -0.25681174 1.46954931]\n",
" [-2.05269934 -0.31497584 -0.37661299]\n",
" [ 1.14121081 -1.09245036 -0.16498684]]\n",
"b2 = \n",
"[[-0.88529978]\n",
" [ 0.03477238]\n",
" [ 0.57537385]]\n",
"v[\"dW1\"] = \n",
"[[-0.11006192 0.11447237 0.09015907]\n",
" [ 0.05024943 0.09008559 -0.06837279]]\n",
"v[\"db1\"] = \n",
"[[-0.01228902]\n",
" [-0.09357694]]\n",
"v[\"dW2\"] = \n",
"[[-0.02678881 0.05303555 -0.06916608]\n",
" [-0.03967535 -0.06871727 -0.08452056]\n",
" [-0.06712461 -0.00126646 -0.11173103]]\n",
"v[\"db2\"] = \n",
"[[ 0.02344157]\n",
" [ 0.16598022]\n",
" [ 0.07420442]]\n",
"s[\"dW1\"] = \n",
"[[ 0.00121136 0.00131039 0.00081287]\n",
" [ 0.0002525 0.00081154 0.00046748]]\n",
"s[\"db1\"] = \n",
"[[ 1.51020075e-05]\n",
" [ 8.75664434e-04]]\n",
"s[\"dW2\"] = \n",
"[[ 7.17640232e-05 2.81276921e-04 4.78394595e-04]\n",
" [ 1.57413361e-04 4.72206320e-04 7.14372576e-04]\n",
" [ 4.50571368e-04 1.60392066e-07 1.24838242e-03]]\n",
"s[\"db2\"] = \n",
"[[ 5.49507194e-05]\n",
" [ 2.75494327e-03]\n",
" [ 5.50629536e-04]]\n",
"```"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"You now have three working optimization algorithms (mini-batch gradient descent, Momentum, Adam). Let's implement a model with each of these optimizers and observe the difference."
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## 5 - Model with different optimization algorithms\n",
"\n",
"Lets use the following \"moons\" dataset to test the different optimization methods. (The dataset is named \"moons\" because the data from each of the two classes looks a bit like a crescent-shaped moon.) "
]
},
{
"cell_type": "code",
"execution_count": 14,
"metadata": {},
"outputs": [
{
"data": {
"image/png": "iVBORw0KGgoAAAANSUhEUgAAAbcAAAD8CAYAAAD0f+rwAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
"text/plain": [
"<matplotlib.figure.Figure at 0x7f113841ca90>"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"train_X, train_Y = load_dataset()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"We have already implemented a 3-layer neural network. You will train it with: \n",
"- Mini-batch **Gradient Descent**: it will call your function:\n",
" - `update_parameters_with_gd()`\n",
"- Mini-batch **Momentum**: it will call your functions:\n",
" - `initialize_velocity()` and `update_parameters_with_momentum()`\n",
"- Mini-batch **Adam**: it will call your functions:\n",
" - `initialize_adam()` and `update_parameters_with_adam()`"
]
},
{
"cell_type": "code",
"execution_count": 15,
"metadata": {
"collapsed": true
},
"outputs": [],
"source": [
"def model(X, Y, layers_dims, optimizer, learning_rate = 0.0007, mini_batch_size = 64, beta = 0.9,\n",
" beta1 = 0.9, beta2 = 0.999, epsilon = 1e-8, num_epochs = 10000, print_cost = True):\n",
" \"\"\"\n",
" 3-layer neural network model which can be run in different optimizer modes.\n",
" \n",
" Arguments:\n",
" X -- input data, of shape (2, number of examples)\n",
" Y -- true \"label\" vector (1 for blue dot / 0 for red dot), of shape (1, number of examples)\n",
" layers_dims -- python list, containing the size of each layer\n",
" learning_rate -- the learning rate, scalar.\n",
" mini_batch_size -- the size of a mini batch\n",
" beta -- Momentum hyperparameter\n",
" beta1 -- Exponential decay hyperparameter for the past gradients estimates \n",
" beta2 -- Exponential decay hyperparameter for the past squared gradients estimates \n",
" epsilon -- hyperparameter preventing division by zero in Adam updates\n",
" num_epochs -- number of epochs\n",
" print_cost -- True to print the cost every 1000 epochs\n",
"\n",
" Returns:\n",
" parameters -- python dictionary containing your updated parameters \n",
" \"\"\"\n",
"\n",
" L = len(layers_dims) # number of layers in the neural networks\n",
" costs = [] # to keep track of the cost\n",
" t = 0 # initializing the counter required for Adam update\n",
" seed = 10 # For grading purposes, so that your \"random\" minibatches are the same as ours\n",
" m = X.shape[1] # number of training examples\n",
" \n",
" # Initialize parameters\n",
" parameters = initialize_parameters(layers_dims)\n",
"\n",
" # Initialize the optimizer\n",
" if optimizer == \"gd\":\n",
" pass # no initialization required for gradient descent\n",
" elif optimizer == \"momentum\":\n",
" v = initialize_velocity(parameters)\n",
" elif optimizer == \"adam\":\n",
" v, s = initialize_adam(parameters)\n",
" \n",
" # Optimization loop\n",
" for i in range(num_epochs):\n",
" \n",
" # Define the random minibatches. We increment the seed to reshuffle differently the dataset after each epoch\n",
" seed = seed + 1\n",
" minibatches = random_mini_batches(X, Y, mini_batch_size, seed)\n",
" cost_total = 0\n",
" \n",
" for minibatch in minibatches:\n",
"\n",
" # Select a minibatch\n",
" (minibatch_X, minibatch_Y) = minibatch\n",
"\n",
" # Forward propagation\n",
" a3, caches = forward_propagation(minibatch_X, parameters)\n",
"\n",
" # Compute cost and add to the cost total\n",
" cost_total += compute_cost(a3, minibatch_Y)\n",
"\n",
" # Backward propagation\n",
" grads = backward_propagation(minibatch_X, minibatch_Y, caches)\n",
"\n",
" # Update parameters\n",
" if optimizer == \"gd\":\n",
" parameters = update_parameters_with_gd(parameters, grads, learning_rate)\n",
" elif optimizer == \"momentum\":\n",
" parameters, v = update_parameters_with_momentum(parameters, grads, v, beta, learning_rate)\n",
" elif optimizer == \"adam\":\n",
" t = t + 1 # Adam counter\n",
" parameters, v, s = update_parameters_with_adam(parameters, grads, v, s,\n",
" t, learning_rate, beta1, beta2, epsilon)\n",
" cost_avg = cost_total / m\n",
" \n",
" # Print the cost every 1000 epoch\n",
" if print_cost and i % 1000 == 0:\n",
" print (\"Cost after epoch %i: %f\" %(i, cost_avg))\n",
" if print_cost and i % 100 == 0:\n",
" costs.append(cost_avg)\n",
" \n",
" # plot the cost\n",
" plt.plot(costs)\n",
" plt.ylabel('cost')\n",
" plt.xlabel('epochs (per 100)')\n",
" plt.title(\"Learning rate = \" + str(learning_rate))\n",
" plt.show()\n",
"\n",
" return parameters"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"You will now run this 3 layer neural network with each of the 3 optimization methods.\n",
"\n",
"### 5.1 - Mini-batch Gradient descent\n",
"\n",
"Run the following code to see how the model does with mini-batch gradient descent."
]
},
{
"cell_type": "code",
"execution_count": 16,
"metadata": {
"scrolled": false
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Cost after epoch 0: 0.702405\n",
"Cost after epoch 1000: 0.668101\n",
"Cost after epoch 2000: 0.635288\n",
"Cost after epoch 3000: 0.600491\n",
"Cost after epoch 4000: 0.573367\n",
"Cost after epoch 5000: 0.551977\n",
"Cost after epoch 6000: 0.532370\n",
"Cost after epoch 7000: 0.514007\n",
"Cost after epoch 8000: 0.496472\n",
"Cost after epoch 9000: 0.468014\n"
]
},
{
"data": {
"image/png": "iVBORw0KGgoAAAANSUhEUgAAAcMAAAEWCAYAAAAadfxCAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n93+o7NbUCU7wmwAAAABJRU5ErkJggg==\n",
"text/plain": [
"<matplotlib.figure.Figure at 0x7f1133446e10>"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"Accuracy: 0.796666666667\n"
]
},
{
"data": {
"image/png": "iVBORw0KGgoAAAANSUhEUgAAAcoAAAEWCAYAAADmYNeIAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nioiIhFCiFBERCfH/50bmP+Yj+ssAAAAASUVORK5CYII=\n",
"text/plain": [
"<matplotlib.figure.Figure at 0x7f1133414940>"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"# train 3-layer model\n",
"layers_dims = [train_X.shape[0], 5, 2, 1]\n",
"parameters = model(train_X, train_Y, layers_dims, optimizer = \"gd\")\n",
"\n",
"# Predict\n",
"predictions = predict(train_X, train_Y, parameters)\n",
"\n",
"# Plot decision boundary\n",
"plt.title(\"Model with Gradient Descent optimization\")\n",
"axes = plt.gca()\n",
"axes.set_xlim([-1.5,2.5])\n",
"axes.set_ylim([-1,1.5])\n",
"plot_decision_boundary(lambda x: predict_dec(parameters, x.T), train_X, train_Y)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### 5.2 - Mini-batch gradient descent with momentum\n",
"\n",
"Run the following code to see how the model does with momentum. Because this example is relatively simple, the gains from using momemtum are small; but for more complex problems you might see bigger gains."
]
},
{
"cell_type": "code",
"execution_count": 17,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Cost after epoch 0: 0.702413\n",
"Cost after epoch 1000: 0.668167\n",
"Cost after epoch 2000: 0.635388\n",
"Cost after epoch 3000: 0.600591\n",
"Cost after epoch 4000: 0.573444\n",
"Cost after epoch 5000: 0.552058\n",
"Cost after epoch 6000: 0.532458\n",
"Cost after epoch 7000: 0.514101\n",
"Cost after epoch 8000: 0.496652\n",
"Cost after epoch 9000: 0.468160\n"
]
},
{
"data": {
"image/png": "iVBORw0KGgoAAAANSUhEUgAAAcMAAAEWCAYAAAAadfxCAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\ne/8fvl8UGzM29/QAAAAASUVORK5CYII=\n",
"text/plain": [
"<matplotlib.figure.Figure at 0x7f11333a4470>"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"Accuracy: 0.796666666667\n"
]
},
{
"data": {
"image/png": "iVBORw0KGgoAAAANSUhEUgAAAcoAAAEWCAYAAADmYNeIAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nYQAAAABJRU5ErkJggg==\n",
"text/plain": [
"<matplotlib.figure.Figure at 0x7f1133463d68>"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"# train 3-layer model\n",
"layers_dims = [train_X.shape[0], 5, 2, 1]\n",
"parameters = model(train_X, train_Y, layers_dims, beta = 0.9, optimizer = \"momentum\")\n",
"\n",
"# Predict\n",
"predictions = predict(train_X, train_Y, parameters)\n",
"\n",
"# Plot decision boundary\n",
"plt.title(\"Model with Momentum optimization\")\n",
"axes = plt.gca()\n",
"axes.set_xlim([-1.5,2.5])\n",
"axes.set_ylim([-1,1.5])\n",
"plot_decision_boundary(lambda x: predict_dec(parameters, x.T), train_X, train_Y)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### 5.3 - Mini-batch with Adam mode\n",
"\n",
"Run the following code to see how the model does with Adam."
]
},
{
"cell_type": "code",
"execution_count": 18,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Cost after epoch 0: 0.702166\n",
"Cost after epoch 1000: 0.167845\n",
"Cost after epoch 2000: 0.141316\n",
"Cost after epoch 3000: 0.138788\n",
"Cost after epoch 4000: 0.136066\n",
"Cost after epoch 5000: 0.134240\n",
"Cost after epoch 6000: 0.131127\n",
"Cost after epoch 7000: 0.130216\n",
"Cost after epoch 8000: 0.129623\n",
"Cost after epoch 9000: 0.129118\n"
]
},
{
"data": {
"image/png": "iVBORw0KGgoAAAANSUhEUgAAAb0AAAEWCAYAAADy9UlpAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAIABJREFUeJzt3XucXHV9//HXe2Z39p5kw26AJEBCCNCggBiCrTcqVsEb\n3otXrFqKSn9q21/FX/1Va2sftdqftRVFvAH1ghdQI0XxUkUUkQQE5CIaIJiESzb3297n8/vjnNlM\nlr0Myc7O7pz38/GYR3bOfOecz5wk897v95zzPYoIzMzMsiBX6wLMzMymi0PPzMwyw6FnZmaZ4dAz\nM7PMcOiZmVlmOPTMzCwzHHpmT4Ck70o6v9Z1mNnBcejZrCBpvaTn1rqOiDgnIq6odR0Akn4i6a01\n2O58Sd+UtFfSQ5JeO0n7d0t6VNIuSZ+X1FTpuiSdJek3kvZJ+rGkY8pe+66kPWWPAUm/nvpPbPXE\noWeWktRQ6xpKZlItY7gEGAAOB14HfErSSWM1lPR84GLgLOAY4FjgHypZl6Qu4Brg/wLzgbXAV0tv\nTH8BaS89gJuAr0/h57R6FBF++DHjH8B64LnjvPYi4HZgB8kX38llr10M3A/sBu4BXlb22puAnwMf\nA7YC/5Qu+xnwUWA78CBwTtl7fgK8tez9E7VdCvw03fYPSb7gvzjOZzgT2Ai8B3gU+C+gE7gW6EnX\nfy2wOG3/IWAY6AP2AJ9Il58I/ADYBtwHvHqK/x7aSELq+LJlVwL/Mk77LwP/XPb8OcCjlawLuAC4\nadS2e4ETx9jOknR/LKn1v1U/ZvbDPT2b1SQ9Bfg88BfAYcCngdVlQ2j3A88E5pL0ML4o6ciyVZwB\nPEDS0/hQ2bL7gC7gX4HPSdI4JUzU9svALWldHwDeMMnHOYKkR3MMyRd+DvhC+vxoki/8TwBExN8B\nNwIXRdLTuUhSG0ngfRlYAJwHfFLSirE2JumTknaM87hznBqPB4Yi4rdly+4AxuzppcvvGNX2cEmH\nVbCuA94bEXuBdeNs643AjRGxfpw6zAAPb9rsdwHw6Yj4ZUQMR3K8rR94GkBEfD0iHo6IYkR8Ffgd\nsKrs/Q9HxH9GxFBE9KbLHoqIz0TEMHAFcCRJKI5lzLaSjgZOB/4+IgYi4mfA6kk+SxF4f0T0R0Rv\nRGyNiKsjYl9E7CYJ5WdP8P4XAesj4gvp5/kVcDXwqrEaR8TbI2LeOI+Tx9lGO7Br1LJdQMcE7XeO\nakvafrJ1jX7vRNt6I3D5ODWYjZjJxw3MKnEMcL6kvyxbVgAWAkh6I/BXJMNfkHyRdpW13TDGOh8t\n/RAR+9KOW/s42x+vbRewLSL2jdrWURN8lp6I6Cs9kdRKMvR6NslQJ0CHpHwasqMdA5whaUfZsgaS\nodKpsgeYM2rZXJIh3Eraz03/3F3BuiralqRnkPSSvzFJ7Wbu6dmstwH40KheSmtEfCU90+8zwEXA\nYRExD7gLKB+qrNZtRh4B5qfBVTJR4I1Vy18DJwBnRMQc4Fnpco3TfgNww6h90R4RbxtrY5IuHXX2\nY/nj7nFq/C3QIGl52bJTgPHa352+Xt72sYjYWsG6DnhvOny7bIxtnQ9cExF7xqnBbIRDz2aTRknN\nZY8GklC7UNIZSrRJeqGkDpITH4LkRBAk/RnwpOkoNCIeIjnb8AOSCpL+EHjxE1xNB8lxvB2S5gPv\nH/X6YyRnQ5ZcCxwv6Q2SGtPH6ZL+YJwaL4yysx9HPcY8RpceV7sG+GC6r58BvITxe5NXAm+RtEJS\nJ8mZmJdXuK5vAk+S9ApJzennvyMiflNauaQW4NV4aNMq5NCz2eQ6khAoPT4QEWuBPyc5wWM7yYkO\nbwKIiHuAfwN+QRIQTyY5W3O6vA74Q/afGfpVkuONlfp3oAXYAtwMfG/U6x8HXilpu6T/SI/7PY/k\nBJaHSYZePww0MbXenta1meSkmbdFxN0Ako5Oe4pHA0TE90hO8Pkx8BDJGa7vr2RdEdEDvILkWOZ2\nkmOx542q5aUkZ+3+eIo/o9UpRfgmsmbTQdJXgd9ExOgem5lNE/f0zKokHVpcJikn6WzgXOBbta7L\nLMt89qZZ9RxBcszqMJILz9+WXkZgZjXi4U0zM8sMD2+amVlmzLrhza6urliyZEmtyzAzsxnk1ltv\n3RIR3ZO1m3Wht2TJEtauXVvrMszMbAaR9FAl7Ty8aWZmmeHQMzOzzHDomZlZZjj0zMwsMxx6ZmaW\nGQ49MzPLjKqGnqSzJd0naZ2ki8d4/X9Luj193CVpOL2FipmZ2ZSrWuhJygOXAOcAK4DXSFpR3iYi\nPhIRp0bEqcB7SW6Aua1aNZV86ZcP8c1fbaz2ZszMbIapZk9vFbAuIh6IiAHgKpJZ5sfzGuArVaxn\nxDdu3cjVt26ajk2ZmdkMUs3QWwRsKHu+MV32OJJagbOBq8d5/QJJayWt7enpOeTCutqb2LLnidzL\n08zM6sFMOZHlxcDPxxvajIjLImJlRKzs7p50arVJOfTMzLKpmqG3CTiq7PnidNlYzmOahjYButsL\nbNs7wHDRt1UyM8uSaobeGmC5pKWSCiTBtnp0I0lzgWcD365iLQfo6miiGLBt78B0bdLMzGaAqoVe\nRAwBFwHXA/cCX4uIuyVdKOnCsqYvA74fEXurVcto3e1NAPTs9hCnmVmWVPXWQhFxHXDdqGWXjnp+\nOXB5NesYrasjCT0f1zMzy5aZciLLtOpqd+iZmWVRRkOvADj0zMyyJpOh197UQFNDji17fCKLmVmW\nZDL0JNHd0eQTWczMMiaToQe+QN3MLIsyHXru6ZmZZUtmQ6+7o+BjemZmGZPZ0Otqb2Lb3n5PRWZm\nliGZDr1iwPZ97u2ZmWVFZkOvu8NTkZmZZU1mQ8+zspiZZU+GQ8+zspiZZU12Q6806fRuH9MzM8uK\nzIZeR1MDhYace3pmZhmS2dCTRLcvUDczy5TMhh4kQ5w97umZmWVGpkOvu92zspiZZUmmQ8+TTpuZ\nZUvmQ2/b3gFPRWZmlhEZD70Cw8XwVGRmZhmR7dDr8KwsZmZZkunQ6273BepmZlmS6dBzT8/MLFuy\nHXqedNrMLFMyHXpzmhso5HO+QN3MLCOqGnqSzpZ0n6R1ki4ep82Zkm6XdLekG6pZzxjbpqu94KnI\nzMwyoqFaK5aUBy4B/gTYCKyRtDoi7ilrMw/4JHB2RPxe0oJq1TOero4mz8piZpYR1ezprQLWRcQD\nETEAXAWcO6rNa4FrIuL3ABGxuYr1jKm7vYkt7umZmWVCNUNvEbCh7PnGdFm544FOST+RdKukN461\nIkkXSForaW1PT8+UFumpyMzMsqPWJ7I0AE8FXgg8H/i/ko4f3SgiLouIlRGxsru7e0oL6OoosHXv\nAEVPRWZmVveqGXqbgKPKni9Ol5XbCFwfEXsjYgvwU+CUKtb0OF3tTZ6KzMwsI6oZemuA5ZKWSioA\n5wGrR7X5NvAMSQ2SWoEzgHurWNPj7L9Wz6FnZlbvqnb2ZkQMSboIuB7IA5+PiLslXZi+fmlE3Cvp\ne8CdQBH4bETcVa2axtJdNivLCXRM56bNzGyaVS30ACLiOuC6UcsuHfX8I8BHqlnHRI6Y0wzAIzv7\nalWCmZlNk1qfyFJzR8xNQu/hHb01rsTMzKot86HX3Jinu6OJTdsdemZm9S7zoQewcF4LD+906JmZ\n1TuHHrB4Xot7emZmGeDQAxbOa2bTjl4ifIG6mVk9c+iRDG/2DxXZttfX6pmZ1TOHHrBoXgsAm3wG\np5lZXXPokfT0wJctmJnVO4cesLgzCb2NPpnFzKyuOfSAuS2NtBbyPLzDs7KYmdUzhx4giYXzWti0\nY1+tSzEzsypy6KUWzWtxT8/MrM459FIL57X4RBYzszrn0Est7mxh694BegeGa12KmZlViUMvtXBe\nercFz8FpZla3HHqpRfNaATwHp5lZHXPopUZ6ej6uZ2ZWtxx6qcPnNJOTQ8/MrJ459FKN+RxHzGlm\no0PPzKxuOfTK+LIFM7P65tArs6izxXdaMDOrYw69MgvntfDozj6Gi76ZrJlZPXLolVk0r4XB4aBn\nd3+tSzEzsypw6JXxzWTNzOqbQ6+MbyZrZlbfHHplSheou6dnZlafqhp6ks6WdJ+kdZIuHuP1MyXt\nlHR7+vj7atYzmY7mRuY0N7inZ2ZWpxqqtWJJeeAS4E+AjcAaSasj4p5RTW+MiBdVq44nalFnq+ff\nNDOrU9Xs6a0C1kXEAxExAFwFnFvF7U2JRfOaPbxpZlanqhl6i4ANZc83pstG+yNJd0r6rqSTxlqR\npAskrZW0tqenpxq1jjh8TrMvWTAzq1O1PpHlNuDoiDgZ+E/gW2M1iojLImJlRKzs7u6uakHz2wps\n3zdA0Reom5nVnWqG3ibgqLLni9NlIyJiV0TsSX++DmiU1FXFmibV2VqgGLCrb7CWZZiZWRVUM/TW\nAMslLZVUAM4DVpc3kHSEJKU/r0rr2VrFmiY1v60AwLa9A7Usw8zMqqBqZ29GxJCki4DrgTzw+Yi4\nW9KF6euXAq8E3iZpCOgFzouImo4rdqaht32fQ8/MrN5ULfRgZMjyulHLLi37+RPAJ6pZwxM1v7XU\n0/PwpplZvan1iSwzTmdbIwDbPbxpZlZ3HHqjjBzT8/CmmVndceiN0tKYp9CQc0/PzKwOOfRGkcT8\n1oLP3jQzq0MOvTF0thXYvs8nspiZ1RuH3hjmtzX6kgUzszrk0BtDZ2vBx/TMzOqQQ28M89sKPnvT\nzKwOOfTG0NlaYGfvIEPDxVqXYmZmU8ihN4b5bQUiYGevT2YxM6snDr0xeP5NM7P65NAbg+ffNDOr\nTw69MZTm3/QF6mZm9cWhN4b5Ht40M6tLDr0xdLb6RrJmZvWootCT9KpKltWL5sY8rYW8L1A3M6sz\nlfb03lvhsrrR2eoL1M3M6s2Ed06XdA7wAmCRpP8oe2kOMFTNwmptfpunIjMzqzcThh7wMLAWeAlw\na9ny3cC7q1XUTNDZVmCb77RgZlZXJgy9iLgDuEPSlyNiEEBSJ3BURGyfjgJrZX5rI+u37K11GWZm\nNoUqPab3A0lzJM0HbgM+I+ljVayr5jo9vGlmVncqDb25EbELeDlwZUScAZxVvbJqr7O1wO7+IQaG\nPOm0mVm9qDT0GiQdCbwauLaK9cwYpfk3d/gMTjOzulFp6H0QuB64PyLWSDoW+F31yqq9kfk3HXpm\nZnVjsrM3AYiIrwNfL3v+APCKahU1E3j+TTOz+lPpjCyLJX1T0ub0cbWkxRW872xJ90laJ+niCdqd\nLmlI0iufSPHVNDL/pu+0YGZWNyod3vwCsBpYmD6+ky4bl6Q8cAlwDrACeI2kFeO0+zDw/crLrj4P\nb5qZ1Z9KQ687Ir4QEUPp43Kge5L3rALWRcQDETEAXAWcO0a7vwSuBjZXWvR0mNda6uk59MzM6kWl\nobdV0usl5dPH64Gtk7xnEbCh7PnGdNkISYuAlwGfmmhFki6QtFbS2p6engpLPjSFhhwdTQ0+pmdm\nVkcqDb03k1yu8CjwCPBK4E1TsP1/B94TERNeDBcRl0XEyohY2d09WQdz6nS2FXzJgplZHano7E2S\nSxbOL009ls7M8lGSMBzPJuCosueL02XlVgJXSQLoAl4gaSgivlVhXVXl+TfNzOpLpaF3cvlcmxGx\nTdJTJnnPGmC5pKUkYXce8NryBhGxtPSzpMuBa2dK4EEy/+aWPe7pmZnVi0qHN3PpRNPASE9vssmq\nh4CLSC5qvxf4WkTcLelCSRcebMHTqbOt4GN6ZmZ1pNKe3r8Bv5BUukD9VcCHJntTRFwHXDdq2aXj\ntH1ThbVMm/mtBbb7mJ6ZWd2odEaWKyWtBZ6TLnp5RNxTvbJmhs62AvsGhukbHKa5MV/rcszM7BBV\n2tMjDbm6D7pyI7Oy7BvgyLktNa7GzMwOVaXH9DKpszQri4/rmZnVBYfeBDz/pplZfXHoTWB+6U4L\nPpnFzKwuOPQmMDK8uae/xpWYmdlUcOhNYF5rgUJDjod39tW6FDMzmwIOvQnkc+LYrjbWbd5T61LM\nzGwKOPQmsay7nft7HHpmZvXAoTeJZQva2bBtH32Dw7UuxczMDpFDbxLLutsoBqzfurfWpZiZ2SFy\n6E3iuAXtANy/2aFnZjbbOfQmcWxXOxI+mcXMrA449CbRUsizaF6LT2YxM6sDDr0KLOtud0/PzKwO\nOPQqcNyCdh7YsodiMWpdipmZHQKHXgWWdbfTN1hk047eWpdiZmaHwKFXgZEzOH1cz8xsVnPoVWBZ\ndxvgMzjNzGY7h14FDmtvorO1kft7fK2emdls5tCr0LLudu53T8/MbFZz6FXouAWeeNrMbLZz6FVo\nWXc7W/cOsH2v76JuZjZbOfQq5DM4zcxmP4dehZZ1J6HnMzjNzGYvh16FFnW20NSQc0/PzGwWq2ro\nSTpb0n2S1km6eIzXz5V0p6TbJa2V9Ixq1nMo8jmxtKvNPT0zs1msaqEnKQ9cApwDrABeI2nFqGY/\nAk6JiFOBNwOfrVY9UyE5g9PX6pmZzVbV7OmtAtZFxAMRMQBcBZxb3iAi9kREaRbnNmBGz+i8rLud\nDdv30TswXOtSzMzsIFQz9BYBG8qeb0yXHUDSyyT9Bvhvkt7e40i6IB3+XNvT01OVYitxylFziYBf\nbdhesxrMzOzg1fxEloj4ZkScCLwU+Mdx2lwWESsjYmV3d/f0FljmqcfMR4I1Dzr0zMxmo2qG3ibg\nqLLni9NlY4qInwLHSuqqYk2HZG5LIyceMYc167fVuhQzMzsI1Qy9NcBySUslFYDzgNXlDSQdJ0np\nz6cBTcDWKtZ0yFYt6eTWh7YzOFysdSlmZvYEVS30ImIIuAi4HrgX+FpE3C3pQkkXps1eAdwl6XaS\nMz3/tOzElhnp9KXz6R0c5u6Hd9W6FDMze4IaqrnyiLgOuG7UskvLfv4w8OFq1jDVVi2ZD8CaB7dx\n6lHzalyNmZk9ETU/kWW2WTCnmSWHtfLLB31cz8xstnHoHYTTl8xn7UPbKBZn9EismZmN4tA7CKuW\nzmfHvkHWeR5OM7NZxaF3EFYtTY7r3eIhTjOzWcWhdxCOnt/Kgo4mh56Z2Szj0DsIkli1dD5r1m9j\nhl9hYWZmZRx6B2nV0vk8srOPjdt7a12KmZlVyKF3kE4vXa/nKcnMzGYNh95BOuHwDuY0N/CL+2f0\nrGlmZlbGoXeQcjnxnBMX8P17HmNgyPNwmpnNBg69Q/DiUxays3eQG39Xu3v8mZlZ5Rx6h+CZy7uZ\n29LI6jsernUpZmZWAYfeISg05HjBk4/gB/c8Ru/AcK3LMTOzSTj0DtGLT1nIvoFhfnjvY7UuxczM\nJuHQO0RnLD2MBR1NHuI0M5sFHHqHKJ8TLzp5ITfc18PO3sFal2NmZhNw6E2Bl5y6kIHhItff9Wit\nSzEzswk49KbAKYvncsxhrR7iNDOb4Rx6U0ASLz55ITfdv4XNu/tqXY6ZmY3DoTdFXn7aIgK48qaH\nal2KmZmNw6E3RY7tbufsk47gil+sZ1efT2gxM5uJHHpT6O1nHsfuviG+eLN7e2ZmM5FDbwo9efFc\nnrm8i8//7EH6Bj1Di5nZTOPQm2Lv+OPj2LJngK+t3VDrUszMbBSH3hQ7Y+l8nnpMJ5++4QEGh33L\nITOzmaSqoSfpbEn3SVon6eIxXn+dpDsl/VrSTZJOqWY900ESbz9zGZt29PLt233dnpnZTFK10JOU\nBy4BzgFWAK+RtGJUsweBZ0fEk4F/BC6rVj3T6TknLuDEIzr4z//5nY/tmZnNINXs6a0C1kXEAxEx\nAFwFnFveICJuiojt6dObgcVVrGfaSOJ9L1zBQ1v38ckfr6t1OWZmlqpm6C0Cys/m2JguG89bgO+O\n9YKkCyStlbS2p2d23KX8Gcu7OPfUhXzqhvtZt3lPrcsxMzNmyIkskv6YJPTeM9brEXFZRKyMiJXd\n3d3TW9wheN8LV9DSmOd93/o1EVHrcszMMq+aobcJOKrs+eJ02QEknQx8Fjg3IrZWsZ5p193RxHvO\nOZGbH9jGNbc97qObmdk0q2borQGWS1oqqQCcB6wubyDpaOAa4A0R8dsq1lIzrzn9aE47eh4fuu5e\ntu8dqHU5ZmaZVrXQi4gh4CLgeuBe4GsRcbekCyVdmDb7e+Aw4JOSbpe0tlr11EouJ/755U9md98g\nb7liDXv6h2pdkplZZmm2HWtauXJlrF07+7Lxe3c9wju+/CtOO3oel//ZKtqaGmpdkplZ3ZB0a0Ss\nnKzdjDiRJQvOftKRfPy8U7n1oe285Yo19A74+j0zs+nm0JtGLzp5IR/701O55cFtvPXKNewb8FCn\nmdl0cuhNs3NPXcRHX3UKv7h/K2/83C3s7PW998zMpotDrwZeftpiPvHa07hj4w5e+5mb2bqnv9Yl\nmZllgkOvRl7w5CO57I0rWbd5D6/+9C94ZGdvrUsyM6t7Dr0a+uMTFnDlm1fx2K5+XvKJn3PLg9tq\nXZKZWV1z6NXYGccextVv+yPamxp47Wdu5gs/f9BTlpmZVYlDbwY44YgOvn3R0znzhAX8w3fu4Z1X\n3c6jO/tqXZaZWd1x6M0Qc5obuewNT+Vvnnc8//3rR3jmv/4Pf/uNO3yHBjOzKeQZWWagDdv28Zkb\nH+CrazYwMFzkWcu7efEpC3neSYczp7mx1uWZmc04lc7I4tCbwbbs6efKm9Zz9W2b2LSjl0I+x7OO\n7+LZJyzg6csOY2lXG5JqXaaZWc059OpIRHD7hh1ce+cjfO+uR9m0I7m84ci5zTzt2MN40qK5PHnR\nXE5aOMdzeppZJjn06lREsH7rPn6+bgs33b+Fteu3s3l3cnG7BIs7W1ja1c6xXW0s7Wrj6PmtLO5s\nYXFnKy2FfI2rNzOrjkpDz92CWUYSS9NAe/3TjgFg864+fr1pJ3dt2sX9PXt4YMsebl2/jb2jJrWe\n19rIgo4mFnQ0093RRGdrgc7WRua1FZjb0khbIU9bUwNthQbamxvoaG6gvamBpoach1HNrC449OrA\ngjnNnDWnmbP+4PCRZRFBz55+NmzrZeP2fWzYto9Hd/XRs7ufnt39rFm/lx37Biu6v19DTjQ15Ghq\nzFPI52gp5GlpzNNayNNSyNOYz5HPica8aMznaGrI0dyYp6khR2M+R0M+R2NONORzNOREPica8iIn\n0ZATuZxGlpeWlYesBE0NOVoak+01N+ZpbsjTXEi2U8jnKORz5HIOZjObmEOvTkliQUczCzqaeeox\nneO2GxgqsqN3gF29g+ztH2Zv/xB7+ofYOzDEnr4hdvcPsbd/iIGhIv1DRfoGh+kbLLJvYIh9A8Ps\n7htiqFhkaDgYLgYDw0X6B4v0Dw3TP1RkcLjI4PD0DKHncxoJ5eaGHM2FJBybGnMjfxby+8O70LA/\nZPO5HA355OfykG7MpyGdhnNe+0M7nzuwXSEN+Hxu/3sO/DNHPi8a0+f5nBD7gzqfF80NyTrMrDoc\nehlXaMiNhGO1RARDxWBwuMhwMQnHoWJQTP8sLRuO/cvKDzUXI0YCt3dgmN7B4SR8h4r0Dw4zMFxk\nIA3YgaEifYNFetO2pfDtGxxm79794T0wVGRguHhADaXwHirW9jh3Y15pSOdpbsyN9JglkVPS8y2F\nbylMG9LQTXrWSY+7YVTvOidRaMjRWtjfS28rNKTPG2hrytPR3Mic5gY6mhtpbvSwttUfh55VnbR/\n6HM2KIX00HAwWNaLLUYcENrDxaQXW2o3mAZpqW3pfcPp+waHk/eMhOyoHvBQcX9g9w3uD+v+wWS9\nyS8CQTE4YBtDxSK9g5H2qosH1D04XGpXpBiMhH0lciIJxaYkFAv5HI0Nyd9jXkICkfxZCtrGkd5x\nEra5NKjLNTYkPeTSsDgA6bryOchJIwGfV9rDziXPS68JiPTvCpJ/Ywf2xpNlI3Wq/HMdOIyefA5G\ntplLl+VGfcZi7P9lbbgY6chA+WdOe/M5kRv1T12UfmHZv+5cuu4g+SUvkt0w8stMMhKQLB9Zj/bX\nHxEESV3ln0tpOyU7Nfn7yZVqVeZ/kXHomY2yP6Shhfo743VwuMi+gaQnXBqm3jeQDG3v6htkd98Q\nu/uG2DeQDHXv6x9m3+Awg6XedBrsyRd1EsJ9g0V29w0xOJyEb3kPuvwM8WIw0usfHC5SjP1f+ETy\nBZ48arZ76l55yAMj4Vv65WDkZ0q/WDASzOVD96VfCqD8F4cD30/Zdsq3n9P+X2ROPGIOl7zutGn7\n/A49s4xpzOeY25JjbsvMnt2nWNZLLoVjKRAj4oCeTBRhOJJeb7HISK+89J6SUs+oFMql9QIHhHgp\nrEe+9CPSHtz+L/zhkfBOh+5HevfFA4bny4NjOE3z8s8yEhZKlpcP95cc0CNMP1upZzrSjU3XV/6L\nROkXjmTkojjS8x/5zJT2EQesv7znWQop2N/bLdVY+myMvHf/eso/e7lIf7EZTms9qrP1EP6VPHEO\nPTObkXI5kSPpcZtNldlxkMXMzGwKOPTMzCwzHHpmZpYZDj0zM8sMh56ZmWVGVUNP0tmS7pO0TtLF\nY7x+oqRfSOqX9DfVrMXMzKxqlyxIygOXAH8CbATWSFodEfeUNdsG/C/gpdWqw8zMrKSaPb1VwLqI\neCAiBoCrgHPLG0TE5ohYAwxWsQ4zMzOguhenLwI2lD3fCJxxMCuSdAFwQfp0j6T7DrE2gC5gyxSs\np155/4zP+2Zi3j8T8/6Z2MHun2MqaTQrZmSJiMuAy6ZynZLWVnKX3azy/hmf983EvH8m5v0zsWrv\nn2oOb24Cjip7vjhdZmZmVhPVDL01wHJJSyUVgPOA1VXcnpmZ2YSqNrwZEUOSLgKuB/LA5yPibkkX\npq9fKukIYC0wByhKehewIiJ2VauuMlM6XFqHvH/G530zMe+fiXn/TKyq+0cx+r4PZmZmdcozspiZ\nWWY49MzMLDMyF3qTTY2WNZKOkvRjSfdIulvSO9Pl8yX9QNLv0j87a11rrUjKS/qVpGvT5943ZSTN\nk/QNSb+RdK+kP/Q+Skh6d/r/6i5JX5HUnOV9I+nzkjZLuqts2bj7Q9J70+/q+yQ9fypqyFTolU2N\ndg6wAnhIFaGqAAAGkklEQVSNpBW1rarmhoC/jogVwNOAd6T75GLgRxGxHPhR+jyr3gncW/bc++ZA\nHwe+FxEnAqeQ7KvM7yNJi0imWVwZEU8iOaHvPLK9by4Hzh61bMz9kX4PnQeclL7nk+l3+CHJVOhR\nwdRoWRMRj0TEbenPu0m+sBaR7Jcr0mZXkNH5USUtBl4IfLZssfdNStJc4FnA5wAiYiAiduB9VNIA\ntEhqAFqBh8nwvomIn5LMuVxuvP1xLnBVRPRHxIPAOpLv8EOStdAba2q0RTWqZcaRtAR4CvBL4PCI\neCR96VHg8BqVVWv/DvwtUCxb5n2z31KgB/hCOgT8WUlteB8REZuAjwK/Bx4BdkbE9/G+GW28/VGV\n7+ushZ6NQ1I7cDXwrtHXSUZyXUvmrm2R9CJgc0TcOl6brO6bMg3AacCnIuIpwF5GDddldR+lx6bO\nJfnFYCHQJun15W2yum/GMx37I2uh56nRxiCpkSTwvhQR16SLH5N0ZPr6kcDmWtVXQ08HXiJpPclQ\n+HMkfRHvm3IbgY0R8cv0+TdIQtD7CJ4LPBgRPRExCFwD/BHeN6ONtz+q8n2dtdDz1GijSBLJ8Zh7\nI+L/lb20Gjg//fl84NvTXVutRcR7I2JxRCwh+bfyPxHxerxvRkTEo8AGSSeki84C7sH7CJJhzadJ\nak3/n51Fcszc++ZA4+2P1cB5kpokLQWWA7cc6sYyNyOLpBeQHKcpTY32oRqXVFOSngHcCPya/cet\n/g/Jcb2vAUcDDwGvjojRB6AzQ9KZwN9ExIskHYb3zQhJp5Kc6FMAHgD+jOQX6szvI0n/APwpyVnS\nvwLeCrST0X0j6SvAmSS3D3oMeD/wLcbZH5L+Dngzyf57V0R895BryFromZlZdmVteNPMzDLMoWdm\nZpnh0DMzs8xw6JmZWWY49MzMLDMcembTTNKZpTs2HOT7Xyrp76eyprJ1f0jSBkl7Ri1vkvTVdMb7\nX6ZT1pVeOz+dIf93ks4vW36VpOXVqNPsYDn0zGafvwU+eagrSSdBHu07jD2p71uA7RFxHPAx4MPp\nOuaTXGt1Rvq+95fdGuZTaa1mM4ZDz2wMkl4v6RZJt0v6dOmWJpL2SPpYeo+0H0nqTpefKulmSXdK\n+mbpi1/ScZJ+KOkOSbdJWpZuor3sHnRfSmfsQNK/KLm34Z2SPjpGXccD/RGxJX1+uaRLJa2V9Nt0\nvtDSPQA/ImlNuq6/SJefKelGSatJZk45QETcXDb5b7nymfC/AZyV1vx84AcRsS0itgM/YP+tY24E\nnjtOuJrVhEPPbBRJf0Ayi8bTI+JUYBh4XfpyG7A2Ik4CbiDp5QBcCbwnIk4mmd2mtPxLwCURcQrJ\nvIulQHkK8C6S+zoeCzw9nenlZcBJ6Xr+aYzyng7cNmrZEpJe1guBSyU1k/TMdkbE6cDpwJ+nUzlB\nMjfmOyPi+CewW0ZmvI+IIWAncBgTzIQfEUWS28Gc8gS2Y1ZVDj2zxzsLeCqwRtLt6fNj09eKwFfT\nn78IPCO9p9y8iLghXX4F8CxJHcCiiPgmQET0RcS+tM0tEbExDYbbSYJrJ9AHfE7Sy4FS23JHktzK\np9zXIqIYEb8jmQbsROB5wBvT+n9JElCl42u3pPcnmw6bSe4wYDYjeNjB7PEEXBER762g7cHO49df\n9vMw0BARQ5JWkYTsK4GLgOeMel8vMHeSGoLkM/xlRFxf/kI6h+jeg6i3NOP9xnS4ci6wNV1+Zlm7\nxcBPyp43pzWbzQju6Zk93o+AV0paAMnJGpKOSV/LkQQSwGuBn0XETmC7pGemy98A3JDeiX6jpJem\n62mS1DreRtN7Gs6NiOuAdzP2sOC9wHGjlr1KUi49XngscB9wPfC29LZRSDo+vbnrwSqfCf+VJHec\niHQ7z5PUmR7HfF66rOR44K5D2K7ZlHJPz2yUiLhH0vuA70vKAYPAO0hmgN8LrEpf30xy7A+SQLg0\nDbXSnQYgCcBPS/pgup5XTbDpDuDb6TE5AX81RpufAv8mSbF/tvjfk9xyZQ5wYUT0SfosyZDpbekJ\nJz3ASyf77JL+lSTMWyVtBD4bER8guf3Uf0laB2wjudUSEbFN0j+S3LYL4INlM+QfDvSmtx8ymxF8\nlwWzJ0DSnohor3ENHwe+ExE/lHQ5cG1EfKOWNY1F0ruBXRHxuVrXYlbi4U2z2eefgXGHSWeQHey/\nzMFsRnBPz8zMMsM9PTMzywyHnpmZZYZDz8zMMsOhZ2ZmmeHQMzOzzPj/lK2+7Y+tk80AAAAASUVO\nRK5CYII=\n",
"text/plain": [
"<matplotlib.figure.Figure at 0x7f11333f2f28>"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"Accuracy: 0.94\n"
]
},
{
"data": {
"image/png": "iVBORw0KGgoAAAANSUhEUgAAAcoAAAEWCAYAAADmYNeIAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nRLUAAAAASUVORK5CYII=\n",
"text/plain": [
"<matplotlib.figure.Figure at 0x7f1132df80f0>"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"# train 3-layer model\n",
"layers_dims = [train_X.shape[0], 5, 2, 1]\n",
"parameters = model(train_X, train_Y, layers_dims, optimizer = \"adam\")\n",
"\n",
"# Predict\n",
"predictions = predict(train_X, train_Y, parameters)\n",
"\n",
"# Plot decision boundary\n",
"plt.title(\"Model with Adam optimization\")\n",
"axes = plt.gca()\n",
"axes.set_xlim([-1.5,2.5])\n",
"axes.set_ylim([-1,1.5])\n",
"plot_decision_boundary(lambda x: predict_dec(parameters, x.T), train_X, train_Y)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### 5.4 - Summary\n",
"\n",
"<table> \n",
" <tr>\n",
" <td>\n",
" **optimization method**\n",
" </td>\n",
" <td>\n",
" **accuracy**\n",
" </td>\n",
" <td>\n",
" **cost shape**\n",
" </td>\n",
"\n",
" </tr>\n",
" <td>\n",
" Gradient descent\n",
" </td>\n",
" <td>\n",
" 79.7%\n",
" </td>\n",
" <td>\n",
" oscillations\n",
" </td>\n",
" <tr>\n",
" <td>\n",
" Momentum\n",
" </td>\n",
" <td>\n",
" 79.7%\n",
" </td>\n",
" <td>\n",
" oscillations\n",
" </td>\n",
" </tr>\n",
" <tr>\n",
" <td>\n",
" Adam\n",
" </td>\n",
" <td>\n",
" 94%\n",
" </td>\n",
" <td>\n",
" smoother\n",
" </td>\n",
" </tr>\n",
"</table> \n",
"\n",
"Momentum usually helps, but given the small learning rate and the simplistic dataset, its impact is almost negligeable. Also, the huge oscillations you see in the cost come from the fact that some minibatches are more difficult thans others for the optimization algorithm.\n",
"\n",
"Adam on the other hand, clearly outperforms mini-batch gradient descent and Momentum. If you run the model for more epochs on this simple dataset, all three methods will lead to very good results. However, you've seen that Adam converges a lot faster.\n",
"\n",
"Some advantages of Adam include:\n",
"- Relatively low memory requirements (though higher than gradient descent and gradient descent with momentum) \n",
"- Usually works well even with little tuning of hyperparameters (except $\\alpha$)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"**References**:\n",
"\n",
"- Adam paper: https://arxiv.org/pdf/1412.6980.pdf"
]
}
],
"metadata": {
"coursera": {
"course_slug": "deep-neural-network",
"graded_item_id": "Ckiv2",
"launcher_item_id": "eNLYh"
},
"kernelspec": {
"display_name": "Python 3",
"language": "python",
"name": "python3"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.6.0"
}
},
"nbformat": 4,
"nbformat_minor": 2
}