Learning Data Mining with Python

 Contents......Page 3
Preface......Page 10
Introducing data mining......Page 15
Using Python and the Jupyter Notebook......Page 17
Installing Python......Page 18
Installing Jupyter Notebook......Page 19
Installing scikit-learn......Page 21
Product recommendations......Page 22
Loading the dataset with NumPy......Page 23
Downloading the example code......Page 25
Implementing a simple ranking of rules......Page 26
Ranking to find the best rules......Page 29
A simple classification example......Page 31
Loading and preparing the dataset......Page 32
Implementing the OneR algorithm......Page 34
Testing the algorithm......Page 36
Summary......Page 39
scikit-learn estimators......Page 40
Nearest neighbors......Page 41
Distance metrics......Page 42
Loading the dataset......Page 45
Moving towards a standard workflow......Page 47
Running the algorithm......Page 48
Setting parameters......Page 49
Preprocessing......Page 51
Standard pre-processing......Page 53
Pipelines......Page 54
Summary......Page 56
Loading the dataset......Page 57
Collecting the data......Page 58
Using pandas to load the dataset......Page 59
Cleaning up the dataset......Page 60
Extracting new features......Page 62
Decision trees......Page 64
Parameters in decision trees......Page 65
Using decision trees......Page 66
Putting it all together......Page 67
Random forests......Page 71
How do ensembles work?......Page 72
Applying random forests......Page 73
Engineering new features......Page 75
Summary......Page 76
Chapter 4: Recommending Movies Using Affinity Analysis......Page 77
Affinity analysis......Page 78
Algorithms for affinity analysis......Page 79
Dealing with the movie recommendation problem......Page 80
Loading with pandas......Page 81
Sparse data formats......Page 82
Understanding the Apriori algorithm and its implementation......Page 83
Looking into the basics of the Apriori algorithm......Page 85
Implementing the Apriori algorithm......Page 86
Extracting association rules......Page 89
Evaluating the association rules......Page 92
Summary......Page 95
Chapter 5: Features and scikit-learn Transformers......Page 96
Representing reality in models......Page 97
Common feature patterns......Page 100
Creating good features......Page 104
Feature selection......Page 105
Selecting the best individual features......Page 107
Feature creation......Page 110
Principal Component Analysis......Page 113
The transformer API......Page 116
Implementing a Transformer......Page 117
Unit testing......Page 118
Putting it all together......Page 119
Summary......Page 120
Chapter 6: Social Media Insight using Naive Bayes......Page 121
Disambiguation......Page 122
Downloading data from a social network......Page 123
Loading and classifying the dataset......Page 125
Creating a replicable dataset from Twitter......Page 129
Bag-of-words models......Page 133
n-gram features......Page 135
Other text features......Page 136
Understanding Bayes' theorem......Page 137
Naive Bayes algorithm......Page 138
How it works......Page 139
Extracting word counts......Page 141
Converting dictionaries to a matrix......Page 142
Putting it all together......Page 143
Evaluation using the F1-score......Page 144
Getting useful features from models......Page 145
Summary......Page 148
Loading the dataset......Page 149
Classifying with an existing model......Page 151
Getting follower information from Twitter......Page 154
Building the network......Page 156
Creating a graph......Page 159
Creating a similarity graph......Page 161
Connected components......Page 165
Optimizing criteria......Page 169
Summary......Page 172
Chapter 8: Beating CAPTCHAs with Neural Networks......Page 174
Artificial neural networks......Page 175
An introduction to neural networks......Page 176
Creating the dataset......Page 178
Drawing basic CAPTCHAs......Page 179
Splitting the image into individual letters......Page 182
Creating a training dataset......Page 185
Training and classifying......Page 187
Back-propagation ......Page 190
Predicting words......Page 191
Ranking mechanisms for word similarity......Page 196
Putting it all together......Page 197
Summary......Page 198
Chapter 9: Authorship Attribution......Page 200
Attributing documents to authors......Page 201
Applications and use cases......Page 202
Authorship attribution......Page 203
Getting the data......Page 205
Using function words......Page 208
Counting function words......Page 209
Classifying with function words......Page 212
Support Vector Machines......Page 213
Classifying with SVMs......Page 214
Character n-grams......Page 215
Extracting character n-grams......Page 216
The Enron dataset......Page 217
Creating a dataset loader......Page 218
Putting it all together......Page 221
Evaluation......Page 222
Summary......Page 224
Chapter 10: Clustering News Articles......Page 226
Using a web API to get data......Page 227
Reddit as a data source......Page 230
Getting the data......Page 231
Finding the stories in arbitrary websites......Page 234
Extracting the content......Page 236
Grouping news articles......Page 238
The k-means algorithm......Page 239
Evaluating the results......Page 242
Extracting topic information from clusters......Page 245
Using clustering algorithms as transformers......Page 246
Evidence accumulation......Page 247
How it works......Page 251
Implementation......Page 253
Online learning......Page 254
Implementation......Page 255
Summary......Page 258
Chapter 11: Object Detection in Images using Deep Neural Networks......Page 259
Use cases......Page 260
Application scenario......Page 262
Intuition......Page 265
Implementing deep neural networks......Page 267
An Introduction to TensorFlow......Page 268
Using Keras......Page 272
Convolutional Neural Networks......Page 277
GPU optimization......Page 279
When to use GPUs for computation......Page 280
Running our code on a GPU......Page 281
Setting up the environment......Page 282
Application......Page 283
Getting the data......Page 284
Creating the neural network......Page 285
Putting it all together......Page 287
Summary......Page 288
Chapter 12: Working with Big Data......Page 290
Big data......Page 291
Applications of big data......Page 292
MapReduce......Page 294
The intuition behind MapReduce......Page 296
A word count example......Page 298
Hadoop MapReduce......Page 300
Getting the data......Page 301
The mrjob package......Page 303
Extracting the blog posts......Page 304
Training Naive Bayes......Page 306
Putting it all together......Page 310
Training on Amazon's EMR infrastructure......Page 315
Summary......Page 319
Scikit-learn tutorials......Page 320
More application ideas......Page 321
More complex pipelines......Page 322
Automated Learning......Page 323
More complex features......Page 324
New datasets......Page 325
Adding noise......Page 326
Spam detection......Page 327
NetworkX......Page 328
Deeper networks......Page 329
Local n-grams......Page 330
Temporal analysis......Page 331
Mahotas......Page 332
Recommendation engine......Page 333
Coursera......Page 334
Index......Page 335