1st Edition

Ensemble Methods Foundations and Algorithms

By Zhi-Hua Zhou Copyright 2012
    236 Pages 73 B/W Illustrations
    by Chapman & Hall

    An up-to-date, self-contained introduction to a state-of-the-art machine learning approach, Ensemble Methods: Foundations and Algorithms shows how these accurate methods are used in real-world tasks. It gives you the necessary groundwork to carry out further research in this evolving field.

    After presenting background and terminology, the book covers the main algorithms and theories, including Boosting, Bagging, Random Forest, averaging and voting schemes, the Stacking method, mixture of experts, and diversity measures. It also discusses multiclass extension, noise tolerance, error-ambiguity and bias-variance decompositions, and recent progress in information theoretic diversity.

    Moving on to more advanced topics, the author explains how to achieve better performance through ensemble pruning and how to generate better clustering results by combining multiple clusterings. In addition, he describes developments of ensemble methods in semi-supervised learning, active learning, cost-sensitive learning, class-imbalance learning, and comprehensibility enhancement.

    Introduction
    Basic Concepts
    Popular Learning Algorithms
    Evaluation and Comparison
    Ensemble Methods
    Applications of Ensemble Methods

    Boosting
    A General Boosting Procedure
    The AdaBoost Algorithm
    Illustrative Examples
    Theoretical Issues
    Multiclass Extension
    Noise Tolerance

    Bagging
    Two Ensemble Paradigms
    The Bagging Algorithm
    Illustrative Examples
    Theoretical Issues
    Random Tree Ensembles

    Combination Methods
    Benefits of Combination
    Averaging
    Voting
    Combining by Learning
    Other Combination Methods
    Relevant Methods

    Diversity
    Ensemble Diversity
    Error Decomposition
    Diversity Measures
    Information Theoretic Diversity
    Diversity Generation

    Ensemble Pruning
    What Is Ensemble Pruning
    Many Could Be Better Than All
    Categorization of Pruning Methods
    Ordering-Based Pruning
    Clustering-Based Pruning
    Optimization-Based Pruning

    Clustering Ensembles
    Clustering
    Categorization of Clustering Ensemble Methods
    Similarity-Based Methods
    Graph-Based Methods
    Relabeling-Based Methods
    Transformation-Based Methods

    Advanced Topics
    Semi-Supervised Learning
    Active Learning
    Cost-Sensitive Learning
    Class-Imbalance Learning
    Improving Comprehensibility
    Future Directions of Ensembles

    References

    Index

    Further Readings appear at the end of each chapter.

    Biography

    Zhi-Hua Zhou is a professor in the Department of Computer Science and Technology and the National Key Laboratory for Novel Software Technology at Nanjing University. Dr. Zhou is the founding steering committee co-chair of ACML and associate editor-in-chief, associate editor, and editorial board member of numerous journals. He has published extensively in top-tier journals, chaired many conferences, and won six international journal/conference/competition awards. His research interests encompass the areas of machine learning, data mining, pattern recognition, and multimedia information retrieval.

    "… a valuable contribution to theoretical and practical ensemble learning. The material is very well presented, preliminaries and basic knowledge are discussed in detail, and many illustrations and pseudo-code tables help to understand the facts of this interesting field of research. Therefore, the book will become a helpful tool for practitioners working in the field of machine learning or pattern recognition as well as for students of engineering or computer sciences at the graduate and postgraduate level. I heartily recommend this book!"
    IEEE Computational Intelligence Magazine, February 2013

    "While the book is rather written for a machine learning and pattern recognition audience, the terminology is well explained and therefore also easily understandable for readers from other areas. In general the book is well structured and written and presents nicely the different ideas and approaches for combining single learners as well as their strengths and limitations."
    —Klaus Nordhausen, International Statistical Review (2013), 81

    "Professor Zhou’s book is a comprehensive introduction to ensemble methods in machine learning. It reviews the latest research in this exciting area. I learned a lot reading it!"
    —Thomas G. Dietterich, Professor and Director of Intelligent Systems Research, Oregon State University, Corvallis, USA; ACM Fellow; and Founding President of the International Machine Learning Society

    "This is a timely book. Right time and right book … with an authoritative but inclusive style that will allow many readers to gain knowledge on the topic."
    —Fabio Roli, University of Cagliari, Italy