The Art of Feature Engineering: Essentials for Machine Learning

by Pablo Duboue

Overview

When machine learning engineers work with data sets, they may find the results aren't as good as they need. Instead of improving the model or collecting more data, they can use the feature engineering process to help improve results by modifying the data's features to better capture the nature of the problem. This practical guide to feature engineering is an essential addition to any data scientist's or machine learning engineer's toolbox, providing new ideas on how to improve the performance of a machine learning solution. Beginning with the basic concepts and techniques, the text builds up to a unique cross-domain approach that spans data on graphs, texts, time series, and images, with fully worked out case studies. Key topics include binning, out-of-fold estimation, feature selection, dimensionality reduction, and encoding variable-length data. The full source code for the case studies is available on a companion website as Python Jupyter notebooks.

Product Details

ISBN-13:	9781108709385
Publisher:	Cambridge University Press
Publication date:	06/25/2020
Pages:	284
Product dimensions:	5.98(w) x 8.98(h) x 0.63(d)

About the Author

Pablo Duboue is Director of Textualization Software Ltd. and is passionate about improving society through technology. He has a Ph.D. in Computer Science from Columbia University and was part of the IBM Watson team that beat the Jeopardy! Champions in 2011. He splits his time between teaching machine learning, doing open research, contributing to free software projects, and consulting for start-ups. He has taught in three different countries and done joint research with more than fifty co-authors. Recent career highlights include a best paper award in the Canadian AI conference industrial track and consulting for a start-up acquired by Intel Corp.

Table of Contents

Part I. Fundamentals: 1. Introduction; 2. Features, combined; 3. Features, expanded; 4. Features, reduced; 5. Advanced topics; Part II. Case Studies: 6. Graph data; 7. Timestamped data; 8. Textual data; 9. Image data; 10. Other domains.

Editorial Reviews

'Pablo Duboue is a true grandmaster of the art and science of feature engineering. His foundational contributions to the creation of IBM Watson were a critical component of its success. Now readers can benefit from his expertise. His book provides deep insights into to how to develop, assess, combine, and enhance machine learning features. Of particular interest to advanced practitioners is his discussion of feature engineering and deep learning; there is a pervasive myth in the industry that deep learning and big data have made feature engineering obsolete, but the book explains why that is often incorrect for real-world computing applications and explains the relationship between building effective features and deep neural network architectures. The book engages with countless other basic and advanced topics in the area of machine learning and feature engineering, making it a valuable resource for machine learning practitioners of all levels of experience.' J. William Murdock, IBM

'Feature engineering is the process of identifying, selecting and evaluating input variables to statistical and machine learning models for a given problem. Pablo Duboue's The Art of Feature Engineering introduces the process with rich detail from a practitioner’s point of view, and adds new insights through four input data scenarios for the same prediction task. Highly recommended!' Nelson Correa, Andinum Inc.

'TAoFE is a comprehensive handbook - sure to be a hit with data science practitioners. With highly accessible and didactic explanations of complex concepts, the book represents the state-of-the-art, and shows in practical terms how it applies to a wide range of real-world case studies.' Gavin Brown, University of Manchester

'This book provides a large catalogue of feature manipulation techniques along with non-trivial examples to illustrate their applicability and impact on performance. It could be suitable as a textbook for an upper level undergrad or graduate text mining or multimodal data analysis class. Recent graduates starting in field data mining and text analysis will find this a useful text.' Wlodek Zadrozny, University of North Carolina

From the Publisher