Cover of Feature Engineering for Machine Learning
DRM-free · Yours to keep forever
Data Science & ML

Feature Engineering for Machine Learning

Crafting, selecting and serving features that move metrics

By Houssam Kodad

PDF 248 pages Intermediate English

One-time purchase

€27.95

VAT included
where applicable

Download sample
  • Instant download after purchase
  • Readable on any device
  • Free updates to this edition
  • Secure checkout

About this book

What's inside

Better features beat fancier models more often than anyone admits. This book is a practical tour of feature engineering for tabular, temporal and text data, with a hard focus on the thing that breaks real systems: training/serving skew. You'll learn to build features that are predictive, reproducible and available at inference time — not just impressive in a notebook.

What you'll learn

Skills you'll walk away with

  • Engineer features for numerical, categorical and date fields
  • Encode high-cardinality categories without leakage
  • Build temporal and window features from event data
  • Avoid target leakage in cross-validation and pipelines
  • Close the gap between training and serving features
  • Select features with importance, permutation and pruning
  • Operate a feature store for online and offline parity

Table of contents

9 chapters
  1. 01

    Why Features Decide the Outcome

    • · Model capacity vs signal
    • · The notebook-to-production gap
    • · A workflow for the book
  2. 02

    Numerical Features Done Right

    • · Scaling and transforms
    • · Binning and outliers
    • · Interactions and ratios
  3. 03

    Encoding Categorical Variables

    • · One-hot vs ordinal
    • · Target and frequency encoding
    • · High-cardinality strategies
  4. 04

    Time, Dates and Window Features

    • · Calendar and cyclical features
    • · Lag and rolling-window features
    • · Avoiding lookahead bias
  5. 05

    Text and Embeddings as Features

    • · Bag-of-words to TF-IDF
    • · Pretrained embeddings
    • · Dimensionality reduction
  6. 06

    The Leakage Traps

    • · Target leakage in practice
    • · Leakage through preprocessing
    • · Leak-proof cross-validation
  7. 07

    Selecting What Matters

    • · Filter, wrapper and embedded methods
    • · Permutation importance
    • · Pruning redundant features
  8. 08

    Training/Serving Parity

    • · Why features drift apart
    • · Shared transformation code
    • · Point-in-time correctness
  9. 09

    Feature Stores in Production

    • · Offline and online stores
    • · Backfills and freshness
    • · Governance and reuse

This is the full chapter list — exactly what you'll receive in the PDF.

More in Data Science & ML

Keep exploring this track

Cover of Practical MLOps: From Notebook to Production
Data Science & ML

Practical MLOps: From Notebook to Production

Packaging, deployment, monitoring and retraining that lasts

€32.95 View details
Cover of Time Series Forecasting in Practice
Data Science & ML

Time Series Forecasting in Practice

Classical models, gradient boosting and deep learning for demand and operations

€26.95 View details
Cover of Recommender Systems at Scale
Data Science & ML

Recommender Systems at Scale

Collaborative filtering, embeddings and multi-stage ranking

€29.95 View details
Cover of Statistical Foundations for Data Scientists
Data Science & ML

Statistical Foundations for Data Scientists

Inference, experimentation and A/B testing done right

€24.95 View details
Cover of Gradient Boosting with XGBoost and LightGBM
Data Science & ML

Gradient Boosting with XGBoost and LightGBM

A practitioner's guide to winning with tabular data

€22.95 View details