{
"cells": [
{
"cell_type": "code",
"execution_count": 1,
"metadata": {
"collapsed": false
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"0.20.3\n",
"0.19.1\n",
"1.14.2\n"
]
}
],
"source": [
"# Imports\n",
"\n",
"import os\n",
"import lime\n",
"import sklearn\n",
"import sklearn.ensemble\n",
"import pandas as pd\n",
"import numpy as np\n",
"from sklearn.datasets import fetch_20newsgroups\n",
"\n",
"print pd.__version__\n",
"print sklearn.__version__\n",
"print np.__version__"
]
},
{
"cell_type": "code",
"execution_count": 2,
"metadata": {
"collapsed": false
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"['alt.atheism', 'comp.graphics', 'comp.os.ms-windows.misc', 'comp.sys.ibm.pc.hardware', 'comp.sys.mac.hardware', 'comp.windows.x', 'misc.forsale', 'rec.autos', 'rec.motorcycles', 'rec.sport.baseball', 'rec.sport.hockey', 'sci.crypt', 'sci.electronics', 'sci.med', 'sci.space', 'soc.religion.christian', 'talk.politics.guns', 'talk.politics.mideast', 'talk.politics.misc', 'talk.religion.misc']\n"
]
}
],
"source": [
"# lists out all the classes possible in the newsgroup dataset\n",
"newsgroups_train = fetch_20newsgroups(subset='train')\n",
"print newsgroups_train.target_names"
]
},
{
"cell_type": "code",
"execution_count": 3,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
"['description', 'DESCR', 'filenames', 'target_names', 'data', 'target']"
]
},
"execution_count": 3,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"# we will be exploring \"comp.sys.ibm.pc.hardware\", \"comp.sys.mac.hardware\"\n",
"categories = [\"comp.sys.ibm.pc.hardware\", \"comp.sys.mac.hardware\"]\n",
"newsgroups_train = fetch_20newsgroups(subset='train', categories=categories)\n",
"newsgroups_test = fetch_20newsgroups(subset='test', categories=categories)\n",
"\n",
"newsgroups_train.keys()"
]
},
{
"cell_type": "code",
"execution_count": 4,
"metadata": {
"collapsed": false
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"From: blakey@ug.cs.dal.ca (Jason \"Fish\" Blakey)\n",
"Subject: Newlife 25 and hard drives\n",
"Nntp-Posting-Host: ug.cs.dal.ca\n",
"Organization: Math, Stats & CS, Dalhousie University, Halifax, NS, Canada\n",
"Lines: 12\n",
"\n",
" Giday netters! Just got a used Newlife 25 accelerator, with FPU, and i \n",
"was wondering about a few points. \n",
"-Anyone know the current driver version for it??\n",
"-Can it handle the 16-bit grayscale card, if i get the video option\n",
"-Why would it be hating my hard drive?(can't use the accelerator and \n",
"\thard drive at the same time). Do i need a new driver on my drive?\n",
"\tWhat make?\n",
"-Thanks,\n",
"\tJason\n",
"-- \n",
" ............................................................................ \n",
" blakey@ug.cs.dal.ca -> He's big! He's purple! He's your best friend!\n",
"\n",
"++++++++++++++++++\n",
"comp.sys.ibm.pc.hardware\n"
]
}
],
"source": [
"# Sample\n",
"print newsgroups_test.get('data')[0]\n",
"print \"++++++++++++++++++\"\n",
"print newsgroups_test.get('target_names')[0]"
]
},
{
"cell_type": "code",
"execution_count": 5,
"metadata": {
"collapsed": false
},
"outputs": [],
"source": [
"# we will be vectorizing the text using TF-IDF vectorization technique\n",
"# we will discuss this TF-IDF in future as part of this challenge itself;\n",
"vectorizer = sklearn.feature_extraction.text.TfidfVectorizer(lowercase=False)\n",
"train_vectors = vectorizer.fit_transform(newsgroups_train.data)\n",
"test_vectors = vectorizer.transform(newsgroups_test.data)"
]
},
{
"cell_type": "code",
"execution_count": 6,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
"((1168, 21486), (777, 21486))"
]
},
"execution_count": 6,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"train_vectors.shape, test_vectors.shape"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Random Forest"
]
},
{
"cell_type": "code",
"execution_count": 7,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
"RandomForestClassifier(bootstrap=True, class_weight=None, criterion='gini',\n",
" max_depth=None, max_features='auto', max_leaf_nodes=None,\n",
" min_impurity_decrease=0.0, min_impurity_split=None,\n",
" min_samples_leaf=1, min_samples_split=2,\n",
" min_weight_fraction_leaf=0.0, n_estimators=500, n_jobs=1,\n",
" oob_score=False, random_state=None, verbose=0,\n",
" warm_start=False)"
]
},
"execution_count": 7,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"# model 1\n",
"rf = sklearn.ensemble.RandomForestClassifier(n_estimators=500)\n",
"rf.fit(train_vectors, newsgroups_train.target)"
]
},
{
"cell_type": "code",
"execution_count": 8,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
"0.8854568854568855"
]
},
"execution_count": 8,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"# do prediction\n",
"pred = rf.predict(test_vectors)\n",
"sklearn.metrics.accuracy_score(newsgroups_test.target, pred)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Naive Bayes"
]
},
{
"cell_type": "code",
"execution_count": 9,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
"MultinomialNB(alpha=1.0, class_prior=None, fit_prior=True)"
]
},
"execution_count": 9,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"# model 2\n",
"from sklearn.naive_bayes import MultinomialNB\n",
"nb = MultinomialNB()\n",
"nb.fit(train_vectors, newsgroups_train.target)"
]
},
{
"cell_type": "code",
"execution_count": 10,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
"0.9124839124839125"
]
},
"execution_count": 10,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"# do prediction\n",
"pred = nb.predict(test_vectors)\n",
"sklearn.metrics.accuracy_score(newsgroups_test.target, pred)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Logistic Classifier"
]
},
{
"cell_type": "code",
"execution_count": 11,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
"LogisticRegression(C=1.0, class_weight=None, dual=False, fit_intercept=True,\n",
" intercept_scaling=1, max_iter=100, multi_class='ovr', n_jobs=1,\n",
" penalty='l2', random_state=None, solver='liblinear', tol=0.0001,\n",
" verbose=0, warm_start=False)"
]
},
"execution_count": 11,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"# model 3\n",
"from sklearn.linear_model import LogisticRegression\n",
"lr = LogisticRegression()\n",
"lr.fit(train_vectors, newsgroups_train.target)"
]
},
{
"cell_type": "code",
"execution_count": 12,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
"0.8867438867438867"
]
},
"execution_count": 12,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"# do prediction\n",
"pred = lr.predict(test_vectors)\n",
"sklearn.metrics.accuracy_score(newsgroups_test.target, pred)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Lime in Action"
]
},
{
"cell_type": "code",
"execution_count": 13,
"metadata": {
"collapsed": true
},
"outputs": [],
"source": [
"from lime import lime_text\n",
"from sklearn.pipeline import make_pipeline\n",
"\n",
"crf = make_pipeline(vectorizer, rf)\n",
"cnb = make_pipeline(vectorizer, nb)\n",
"clr = make_pipeline(vectorizer, lr)"
]
},
{
"cell_type": "code",
"execution_count": 14,
"metadata": {
"collapsed": true
},
"outputs": [],
"source": [
"from lime.lime_text import LimeTextExplainer\n",
"explainer = LimeTextExplainer(class_names=['ibm', 'mac'])"
]
},
{
"cell_type": "code",
"execution_count": 15,
"metadata": {
"collapsed": false
},
"outputs": [],
"source": [
"# picking on random example from the test dataset; and seeing the top 6 features learnt by each different\n",
"# classifier for predicting the actual class of the example data point\n",
"idx = np.random.randint(1, len(newsgroups_test.data))\n",
"exp_crf = explainer.explain_instance(newsgroups_test.data[idx], crf.predict_proba, num_features=6)\n",
"exp_clr = explainer.explain_instance(newsgroups_test.data[idx], clr.predict_proba, num_features=6)\n",
"exp_cnb = explainer.explain_instance(newsgroups_test.data[idx], cnb.predict_proba, num_features=6)"
]
},
{
"cell_type": "code",
"execution_count": 16,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/html": [
"\n",
" \n",
" \n",
" \n",
" \n",
" \n",
" "
],
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"exp_crf.show_in_notebook(text=True)"
]
},
{
"cell_type": "code",
"execution_count": 17,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/html": [
"\n",
" \n",
" \n",
" \n",
" \n",
" \n",
" "
],
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"exp_cnb.show_in_notebook(text=True)"
]
},
{
"cell_type": "code",
"execution_count": 18,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/html": [
"\n",
" \n",
" \n",
" \n",
" \n",
" \n",
" "
],
"text/plain": [
""
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"exp_clr.show_in_notebook(text=True)"
]
}
],
"metadata": {
"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": "2.7.12"
}
},
"nbformat": 4,
"nbformat_minor": 1
}