Python Machine Learning By Example

Table of contents

1. Credits
2. Preface
   1. What this book covers
   2. What you need for this book
   3. Who this book is for
   4. Conventions
   5. Reader feedback
   6. Customer support
      1. Downloading the example code
      2. Errata
      3. Piracy
      4. Questions
3. Getting Started with Python and Machine Learning
   1. What is machine learning and why do we need it?
   2. A very high level overview of machine learning
   3. A brief history of the development of machine learning algorithms
   4. Generalizing with data
   5. Overfitting, underfitting and the bias-variance tradeoff
      1. Avoid overfitting with cross-validation
      2. Avoid overfitting with regularization
   6. Avoid overfitting with feature selection and dimensionality reduction
   7. Preprocessing, exploration, and feature engineering
      1. Missing values
      2. Label encoding
      3. One-hot-encoding
      4. Scaling
      5. Polynomial features
      6. Power transformations
      7. Binning
   8. Combining models
      1. Bagging
      2. Boosting
      3. Stacking
      4. Blending
      5. Voting and averaging
   9. Installing software and setting up
   10. Troubleshooting and asking for help
   11. Summary
4. Exploring the 20 Newsgroups Dataset with Text Analysis Algorithms
   1. What is NLP?
   2. Touring powerful NLP libraries in Python
   3. The newsgroups data
   4. Getting the data
   5. Thinking about features
   6. Visualization
   7. Data preprocessing
   8. Clustering
   9. Topic modeling
   10. Summary
5. Spam Email Detection with Naive Bayes
   1. Getting started with classification
   2. Types of classification
   3. Applications of text classification
   4. Exploring naive Bayes
   5. Bayes' theorem by examples
   6. The mechanics of naive Bayes
   7. The naive Bayes implementations
   8. Classifier performance evaluation
   9. Model tuning and cross-validation
   10. Summary
6. News Topic Classification with Support Vector Machine
   1. Recap and inverse document frequency
   2. Support vector machine
      1. The mechanics of SVM
         1. Scenario 1 - identifying the separating hyperplane
         2. Scenario 2 - determining the optimal hyperplane
         3. Scenario 3 - handling outliers
      2. The implementations of SVM
         1. Scenario 4 - dealing with more than two classes
      3. The kernels of SVM
      4. Choosing between the linear and RBF kernel
   3. News topic classification with support vector machine
   4. More examples - fetal state classification on cardiotocography with SVM
   5. Summary
7. Click-Through Prediction with Tree-Based Algorithms
   1. Brief overview of advertising click-through prediction
   2. Getting started with two types of data, numerical and categorical
   3. Decision tree classifier
      1. The construction of a decision tree
      2. The metrics to measure a split
      3. The implementations of decision tree
   4. Click-through prediction with decision tree
   5. Random forest - feature bagging of decision tree
   6. Summary
8. Click-Through Prediction with Logistic Regression
   1. One-hot encoding - converting categorical features to numerical
   2. Logistic regression classifier
      1. Getting started with the logistic function
      2. The mechanics of logistic regression
      3. Training a logistic regression model via gradient descent
   3. Click-through prediction with logistic regression by gradient descent
      1. Training a logistic regression model via stochastic gradient descent
      2. Training a logistic regression model with regularization
      3. Training on large-scale datasets with online learning
      4. Handling multiclass classification
   4. Feature selection via random forest
   5. Summary
9. Stock Price Prediction with Regression Algorithms
   1. Brief overview of the stock market and stock price
   2. What is regression?
   3. Predicting stock price with regression algorithms
      1. Feature engineering
      2. Data acquisition and feature generation
      3. Linear regression
      4. Decision tree regression
      5. Support vector regression
      6. Regression performance evaluation
      7. Stock price prediction with regression algorithms
   4. Summary
10. Best Practices
    1. Machine learning workflow
    2. Best practices in the data preparation stage
       1. Best practice 1 - completely understand the project goal
       2. Best practice 2 - collect all fields that are relevant
       3. Best practice 3 - maintain consistency of field values
       4. Best practice 4 - deal with missing data
    3. Best practices in the training sets generation stage
       1. Best practice 5 - determine categorical features with numerical values
       2. Best practice 6 - decide on whether or not to encode categorical features
       3. Best practice 7 - decide on whether or not to select features and if so, how
       4. Best practice 8 - decide on whether or not to reduce dimensionality and if so how
       5. Best practice 9 - decide on whether or not to scale features
       6. Best practice 10 - perform feature engineering with domain expertise
       7. Best practice 11 - perform feature engineering without domain expertise
       8. Best practice 12 - document how each feature is generated
    4. Best practices in the model training, evaluation, and selection stage
       1. Best practice 13 - choose the right algorithm(s) to start with
          1. Naive Bayes
          2. Logistic regression
          3. SVM
          4. Random forest (or decision tree)
          5. Neural networks
       2. Best practice 14 - reduce overfitting
       3. Best practice 15 - diagnose overfitting and underfitting
    5. Best practices in the deployment and monitoring stage
       1. Best practice 16 - save, load, and reuse models
       2. Best practice 17 - monitor model performance
       3. Best practice 18 - update models regularly
    6. Summary