Feature Engineering for Machine Learning. Principles and Techniques for Data Scientists

Spis treści ebooka

• Preface
  • Introduction
  • Conventions Used in This Book
  • Using Code Examples
  • OReilly Safari
  • How to Contact Us
  • Acknowledgments
    • Special Thanks from Alice
    • Special Thanks from Amanda
• 1. The Machine Learning Pipeline
  • Data
  • Tasks
  • Models
  • Features
  • Model Evaluation
• 2. Fancy Tricks with Simple Numbers
  • Scalars, Vectors, and Spaces
  • Dealing with Counts
    • Binarization
    • Quantization or Binning
      • Fixed-width binning
      • Quantile binning
  • Log Transformation
    • Log Transform in Action
    • Power Transforms: Generalization of the Log Transform
  • Feature Scaling or Normalization
    • Min-Max Scaling
    • Standardization (Variance Scaling)
    • 2 Normalization
  • Interaction Features
  • Feature Selection
  • Summary
  • Bibliography
• 3. Text Data: Flattening, Filtering, and Chunking
  • Bag-of-X: Turning Natural Text into Flat Vectors
    • Bag-of-Words
    • Bag-of-n-Grams
  • Filtering for Cleaner Features
    • Stopwords
    • Frequency-Based Filtering
      • Frequent words
      • Rare words
    • Stemming
  • Atoms of Meaning: From Words to n-Grams to Phrases
    • Parsing and Tokenization
    • Collocation Extraction for Phrase Detection
      • Frequency-based methods
      • Hypothesis testing for collocation extraction
      • Chunking and part-of-speech tagging
  • Summary
  • Bibliography
• 4. The Effects of Feature Scaling: From Bag-of-Words to Tf-Idf
  • Tf-Idf : A Simple Twist on Bag-of-Words
  • Putting It to the Test
    • Creating a Classification Dataset
    • Scaling Bag-of-Words with Tf-Idf Transformation
    • Classification with Logistic Regression
    • Tuning Logistic Regression with Regularization
  • Deep Dive: What Is Happening?
  • Summary
  • Bibliography
• 5. Categorical Variables: Counting Eggs in the Age of Robotic Chickens
  • Encoding Categorical Variables
    • One-Hot Encoding
    • Dummy Coding
    • Effect Coding
    • Pros and Cons of Categorical Variable Encodings
  • Dealing with Large Categorical Variables
    • Feature Hashing
    • Bin Counting
      • What about rare categories?
      • Guarding against data leakage
      • Counts without bounds
  • Summary
  • Bibliography
• 6. Dimensionality Reduction: Squashing the Data Pancake with PCA
  • Intuition
  • Derivation
    • Linear Projection
    • Variance and Empirical Variance
    • Principal Components: First Formulation
    • Principal Components: Matrix-Vector Formulation
    • General Solution of the Principal Components
    • Transforming Features
    • Implementing PCA
  • PCA in Action
  • Whitening and ZCA
  • Considerations and Limitations of PCA
  • Use Cases
  • Summary
  • Bibliography
• 7. Nonlinear Featurization via K-Means Model Stacking
  • k-Means Clustering
  • Clustering as Surface Tiling
  • k-Means Featurization for Classification
    • Alternative Dense Featurization
  • Pros, Cons, and Gotchas
  • Summary
  • Bibliography
• 8. Automating the Featurizer: Image Feature Extraction and Deep Learning
  • The Simplest Image Features (and Why They Dont Work)
  • Manual Feature Extraction: SIFT and HOG
    • Image Gradients
    • Gradient Orientation Histograms
      • How many bins should there be? Should they span from 0°360° (signed gradients) or 0°180° (unsigned gradients)?
      • What weight functions should be used?
      • How are neighborhoods defined? How should they cover the image?
      • What kind of normalization should be done?
    • SIFT Architecture
  • Learning Image Features with Deep Neural Networks
    • Fully Connected Layers
    • Convolutional Layers
    • Rectified Linear Unit (ReLU) Transformation
    • Response Normalization Layers
    • Pooling Layers
    • Structure of AlexNet
  • Summary
  • Bibliography
• 9. Back to the Feature: Building an Academic Paper Recommender
  • Item-Based Collaborative Filtering
  • First Pass: Data Import, Cleaning, and Feature Parsing
    • Academic Paper Recommender: Naive Approach
  • Second Pass: More Engineering and a Smarter Model
    • Academic Paper Recommender: Take 2
  • Third Pass: More Features = More Information
    • Academic Paper Recommender: Take 3
  • Summary
  • Bibliography
• A. Linear Modeling and Linear Algebra Basics
  • Overview of Linear Classification
  • The Anatomy of a Matrix
    • From Vectors to Subspaces
    • Singular Value Decomposition (SVD)
    • The Four Fundamental Subspaces of the Data Matrix
      • Column space
      • Row space
      • Null space
      • Left null space
  • Solving a Linear System
  • Bibliography
• Index

pokaż cały spis treści

ukryj recenzje