附註:Includes bibliographical references and index.
Invited Papers -- Classifier Ensembles for Changing Environments -- A Generic Sensor Fusion Problem: Classification and Function Estimation -- Bagging and Boosting -- AveBoost2: Boosting for Noisy Data -- Bagging Decision Multi-trees -- Learn++. MT: A New Approach to Incremental Learning -- Beyond Boosting: Recursive ECOC Learning Machines -- Exact Bagging with k-Nearest Neighbour Classifiers -- Combination Methods -- Yet Another Method for Combining Classifiers Outputs: A Maximum Entropy Approach -- Combining One-Class Classifiers to Classify Missing Data -- Combining Kernel Information for Support Vector Classification -- Combining Classifiers Using Dependency-Based Product Approximation with Bayes Error Rate -- Combining Dissimilarity-Based One-Class Classifiers -- A Modular System for the Classification of Time Series Data -- A Probabilistic Model Using Information Theoretic Measures for Cluster Ensembles -- Classifier Fusion Using Triangular Norms -- Dynamic Integration of Regression Models -- Dynamic Classifier Selection by Adaptive k-Nearest-Neighbourhood Rule -- Design Methods -- Spectral Measure for Multi-class Problems -- The Relationship between Classifier Factorisation and Performance in Stochastic Vector Quantisation -- A Method for Designing Cost-Sensitive ECOC -- Building Graph-Based Classifier Ensembles by Random Node Selection -- A Comparison of Ensemble Creation Techniques -- Multiple Classifiers System for Reducing Influences of Atypical Observations -- Sharing Training Patterns among Multiple Classifiers -- Performance Analysis -- First Experiments on Ensembles of Radial Basis Functions -- Random Aggregated and Bagged Ensembles of SVMs: An Empirical Bias-Variance Analysis -- Building Diverse Classifier Outputs to Evaluate the Behavior of Combination Methods: The Case of Two Classifiers -- An Empirical Comparison of Hierarchical vs. Two-Level Approaches to Multiclass Problems -- Experiments on Ensembles with Missing and Noisy Data -- Applicatio
摘要:This book constitutes the refereed proceedings of the 5th International Workshop on Multiple Classifier Systems, MCS 2004, held in Cagliari, Italy in June 2004. The 35 revised full papers presented together with 2 invited papers were carefully reviewed and selected from 50 submissions. The papers are organized in topical sections on bagging and boosting, combination methods, design methods, performance analysis, and applications.
