Classification, Clustering, and Data Analysis: Recent Advances and Applications (Studies in Classification, Data Analysis, and Knowledge Organization)

The present volume contains a selection of papers presented at the Eighth Conference of the International Federation of Classification Societies (IFCS) which was held in Cracow, Poland, July 16-19, 2002. All originally submitted papers were subject to a reviewing process by two independent referees, a procedure which resulted in the selection of the 53 articles presented in this volume. These articles relate to theoretical investigations as well as to practical applications and cover a wide range of topics in the broad domain of classifi­ cation, data analysis and related methods. If we try to classify the wealth of problems, methods and approaches into some representative (partially over­ lapping) groups, we find in particular the following areas: • Clustering • Cluster validation • Discrimination • Multivariate data analysis • Statistical methods • Symbolic data analysis • Consensus trees and phylogeny • Regression trees • Neural networks and genetic algorithms • Applications in economics, medicine, biology, and psychology. Given the international orientation of IFCS conferences and the leading role of IFCS in the scientific world of classification, clustering and data anal­ ysis, this volume collects a representative selection of current research and modern applications in this field and serves as an up-to-date information source for statisticians, data analysts, data mining specialists and computer scientists. Front Matter....Pages I-XI Front Matter....Pages 1-1 Some Thoughts about Classification....Pages 5-26 Partial Defuzzification of Fuzzy Clusters....Pages 27-33 A New Clustering Approach, Based on the Estimation of the Probability Density Function, for Gene Expression Data....Pages 35-42 Two-mode Partitioning: Review of Methods and Application of Tabu Search....Pages 43-51 Dynamical Clustering of Interval Data: Optimization of an Adequacy Criterion Based on Hausdorff Distance....Pages 53-60 Removing Separation Conditions in a 1 against 3-Components Gaussian Mixture Problem....Pages 61-73 Obtaining Partitions of a Set of Hard or Fuzzy Partitions....Pages 75-79 Clustering for Prototype Selection using Singular Value Decomposition....Pages 81-88 Clustering in High-dimensional Data Spaces....Pages 89-96 Quantization of Models: Local Approach and Asymptotically Optimal Partitions....Pages 97-105 The Performance of an Autonomous Clustering Technique....Pages 107-112 Cluster Analysis by Restricted Random Walks....Pages 113-120 Missing Data in Hierarchical Classification of Variables — a Simulation Study....Pages 121-128 Representation and Evaluation of Partitions....Pages 131-138 Assessing the Number of Clusters of the Latent Class Model....Pages 139-146 Validation of Very Large Data Sets Clustering by Means of a Nonparametric Linear Criterion....Pages 147-157 Effect of Feature Selection on Bagging Classifiers Based on Kernel Density Estimators....Pages 161-168 Biplot Methodology for Discriminant Analysis Based upon Robust Methods and Principal Curves....Pages 169-176 Bagging Combined Classifiers....Pages 177-184 Application of Bayesian Decision Theory to Constrained Classification Networks....Pages 185-190 Front Matter....Pages 191-191 Quotient Dissimilarities, Euclidean Embeddability, and Huygens’ Weak Principle....Pages 195-202 Conjoint Analysis and Stimulus Presentation — a Comparison of Alternative Methods....Pages 203-210 Grade Correspondence-cluster Analysis Applied to Separate Components of Reversely Regular Mixtures....Pages 211-218 Obtaining Reducts with a Genetic Algorithm....Pages 219-225 A Projection Algorithm for Regression with Collinearity....Pages 227-234 Confronting Data Analysis with Constructivist Philosophy....Pages 235-243 Maximum Likelihood Clustering with Outliers....Pages 247-255 An Improved Method for Estimating the Modes of the Probability Density Function and the Number of Classes for PDF-based Clustering....Pages 257-262 Maximization of Measure of Allowable Sample Sizes Region in Stratified Sampling....Pages 263-269 On Estimation of Population Averages on the Basis of Cluster Sample....Pages 271-277 Symbolic Regression Analysis....Pages 281-288 Modelling Memory Requirement with Normal Symbolic Form....Pages 289-296 Mixture Decomposition of Distributions by Copulas....Pages 297-310 Determination of the Number of Clusters for Symbolic Objects Described by Interval Variables....Pages 311-318 Symbolic Data Analysis Approach to Clustering Large Datasets....Pages 319-327 Symbolic Class Descriptions....Pages 329-337 A Comparison of Alternative Methods for Detecting Reticulation Events in Phylogenetic Analysis....Pages 341-347 Hierarchical Clustering of Multiple Decision Trees....Pages 349-357 Multiple Consensus Trees....Pages 359-364 A Family of Average Consensus Methods for Weighted Trees....Pages 365-369 Front Matter....Pages 191-191 Comparison of Four Methods for Inferring Additive Trees from Incomplete Dissimilarity Matrices....Pages 371-378 Quartet Trees as a Tool to Reconstruct Large Trees from Sequences....Pages 379-388 Regression Trees for Longitudinal Data with Time-Dependent Covariates....Pages 391-398 Tree-based Models in Statistics: Three Decades of Research....Pages 399-407 Computationally Efficient Linear Regression Trees....Pages 409-415 A Clustering Based Procedure for Learning the Hidden Unit Parameters in Elliptical Basis Function Networks....Pages 419-426 Multi-layer Perceptron on Interval Data....Pages 427-434 Front Matter....Pages 435-435 Textual Analysis of Customer Statements for Quality Control and Help Desk Support....Pages 437-445 AHP as Support for Strategy Decision Making in Banking....Pages 447-453 Bioinformatics and Classification: The Analysis of Genome Expression Data....Pages 455-461 Glaucoma Diagnosis by Indirect Classifiers....Pages 463-470 A Cluster Analysis of the Importance of Country and Sector on Company Returns....Pages 471-477 Problems of Classification in Investigative Psychology....Pages 479-487 Back Matter....Pages 489-492

This book deals with recent developments in classification and data analysis and presents new topics which are of central interest to modern statistics. In particular, these include: classification models and clustering methods, multivariate data analysis, symbolic data, neural networks and learning devices, phylogeny and bioinformatics, new software systems for classification and data analysis, as well as applications in social, economic, biological, medical and other sciences. The book presents a long list of useful methods for classification, clustering and data analysis. By combining theoretical aspects with practical problems it is designed for researchers as well as for applied statisticians and will support the fast transfer of new methodological advances to a wide range of applications.

Abstract. The paper contains some general remarks on the high art of data analysis, some philosophical thoughts about classification, a partial review of outliers and robustness from the point of view of applications, including a discussion of the problem of model choice, and a review of several aspects of robust estimation of covariance matrices, including the pragmatic choice of a weight function based on empirical and theoretical evidence. Clustering and Discrimination: Clustering; Cluster Validation; Discrimination Multivariate Data Analysis and Statistics: Multivariate Data Analysis; Statistical Methods; Symbolic Data Analysis; Consensus Trees and Phylogenetics; Regression Trees; Neural Networks and Genetic Algorithms Applications. Dealing with developments in classification and data analysis, this book presents topics of interest to modern statistics, including classification models and clustering methods, multi-variate data analysis, symbolic data, neural networks and learning devices, phylogeny, bioinformatics and more