Hayashi / Jambu / Diday Recent Developments in Clustering and Data Analysis

Développements Récents en Classification Automatique et Analyse des Données: Proceedings of the Japanese-French Scientific Seminar March 24-26, 1987
1. Auflage 2014
ISBN: 978-1-4832-6309-0
Verlag: Elsevier Science & Techn.
Format: PDF
Kopierschutz: 1 - PDF Watermark

E-Book, Englisch, 468 Seiten, Web PDF

Recent Developments in Clustering and Data Analysis presents the results of clustering and multidimensional data analysis research conducted primarily in Japan and France. This book focuses on the significance of the data itself and on the informatics of the data. Organized into four sections encompassing 35 chapters, this book begins with an overview of the quantification of qualitative data as a method of analyzing statistically multidimensional data. This text then examines the rules of interpretation of correspondence cluster analysis by selecting classes and explaining variables involved in the algorithm of hierarchical classification. Other chapters consider the bootstrap and cross-validation methods, which are applied to the logistic ad nonparametric regression analyses of ordered categorical responses. The final chapter deals with a simpler treatment to classify the sleep state. This book is a valuable resource for researchers and workers in the fields from the behavioral sciences, biological sciences, medicine, and industrial sciences.

Autoren/Hrsg.

Hayashi, Chikio

Jambu, Michel

Diday, Edwin

Inhaltsverzeichnis

1;Front Cover;1
2;Recent Developments in Clustering and Data Analysis: Développements Récents en Classification Automatique et Analyse des Données;4
3;Copyright Page;5
4;Table of Contents;6
5;Contributors;10
6;Preface;14
7;Section 1: Data Analysis Techniques and Related Topics with Statistical Software;18
7.1;CHAPTER 1. NEW DEVELOPMENTS IN MULTIDIMENSIONAL DATA ANALYSIS;20
7.1.1;I. INTRODUCTION;20
7.1.2;II. BIRTH OF QUANTIFICATION OF QUALITATIVE DATA;21
7.1.3;III. OVERVIEW OF QUANTIFICATION METHODS;28
7.1.4;IV. FUTURE PROBLEMS;31
7.1.5;REFERENCES;32
7.2;CHAPTER 2. INTERPRETATION OF SOME DATA ANALYSIS METHODS;34
7.2.1;I. INTRODUCTION;34
7.2.2;II. CORRESPONDENCE CLUSTER ANALYSIS;34
7.2.3;III. RESULTS;46
7.2.4;V. CONCLUSION;51
7.2.5;REFERENCES;52
7.3;CHAPTER 3.
A GENERAL EUCLIDEAN APPROACH FOR MEASURING AND DESCRIBING ASSOCIATIONS BETWEEN SEVERAL SETS OF VARIABLES;54
7.3.1;I. INTRODUCTION;54
7.3.2;II. SOME USEFUL DISTANCES;55
7.3.3;III. GENERAL ASSOCIATION INDICES;59
7.3.4;IV. SYNTHETIC RELATIONSHIP METHOD;62
7.3.5;V. SYNTHETIC CONCLUSIONS;63
7.3.6;REFERENCES;64
7.4;CHAPTER 4.
DATA-ADAPTIVE METHODS IN MULTIVARIATE ANALYSIS;66
7.4.1;I. INTRODUCTION;66
7.4.2;II. DATA AND MODEL;67
7.4.3;III. NONPARAMETRIC REGRESSION;70
7.4.4;IV. LOGISTIC REGRESSION ANALYSIS;73
7.4.5;V. RESIDUAL ANALYSIS;76
7.4.6;VI. CONCLUDING REMARKS;79
7.4.7;ACKNOWLEDGEMENTS;81
7.4.8;REFERENCES;81
7.5;CHAPTER 5.
SPECIFIC DESIGN OF A SOFTWARE FOR MULTIVARIATE DESCRIPTIVE STATISTICAL ANALYSIS THE CASE OF SPAD.N;84
7.5.1;I. GENERAL VIEW OF THE SOFTWARE;84
7.5.2;II. DESIGN OF THE SOFTWARE: MODULARITY;85
7.5.3;III. EASY SELECTION OF RELEVANT DATA (PROC SELEC);87
7.5.4;IV. SAVINGS IN THE COMPUTATIONAL STEP (PROC CORMU);88
7.5.5;V. SPECIFIC TOOLS FOR CLUSTERING (PROC RECIP);90
7.5.6;VI. AUTOMATIC CHARACTERIZATION OF CLASSES (PROC DECLA);91
7.5.7;VII. THE GRAPHICAL TOOLS (PROC GRAPH);92
7.5.8;VIII. CREATING AND COPYING VARIABLES (PROC ESCAL);93
7.5.9;IX. EXTENSIONS OF THE SOFTWARE;93
7.5.10;REFERENCES;94
7.6;CHAPTER 6.
A TEST OF GOODNESS OF FIT BASED ON GRAPHICAL METHOD;96
7.6.1;I. INTRODUCTION;96
7.6.2;II. GRAPHICAL REPRESENTATION AND A TEST STATISTIC;97
7.6.3;III. CALCULATION OF THE EXACT PERCENT POINTS;99
7.6.4;IV. COMPARISONS OF THE POWER AMONG T1, T2, T3, W AND Lw;101
7.6.5;REFERENCES;102
7.7;CHAPTER 7.
GRAPHICAL ANALYSIS OF RANKS;104
7.7.1;I. INTRODUCTION;104
7.7.2;II. GRAPHICAL REPRESENTATION OF RANKS;104
7.7.3;III. GRAPHICAL TEST OF RANKS;107
7.7.4;IV. EXAMPLE;108
7.7.5;V. DISCUSSION;110
7.7.6;SUMMARY;112
7.7.7;APPENDIX;112
7.7.8;REFERENCES;113
7.8;CHAPTER 8.
A UNIFIED STUDY OF MULTIVARIATE DATA ANALYSIS METHODS BY NONLINEAR FORMULATIONS AND UNDERLYING PROBABILISTIC STRUCTURES;114
7.8.1;I. INTRODUCTION;114
7.8.2;II. DEFINITIONS AND BASIC CONCEPTS;115
7.8.3;III. NONLINEAR EXTENSIONS OF LINEAR METHODS;116
7.8.4;IV. INTERPRETATION OF QUANTIFICATION METHODS;118
7.8.5;REFERENCES;119
7.9;CHAPTER 9.
OPTIMUM CLASSIFICATION BOUNDARIES BASED ON A CONCOMITANT VARIABLE IN SAMPLE SURVEY SOME APPLICATION TO THE CURRENT STATISTICS OF COMMERCE;120
7.9.1;I. INTRODUCTION;120
7.9.2;II. OPTIMUM STRATIFICATION BASED ON A CONCOMITANT VARIABLE;121
7.9.3;III. ROBUSTNESS ON A REGRESSION FUNCTION AND THE CONSTANT C;122
7.9.4;IV. SOME NUMERICAL EXAMPLE;123
7.9.5;REFERENCES;125
7.10;CHAPTER 10.
ON THE IDENTIFICATION PROBLEM OF AGE-PERIOD-COHORT ANALYSIS;126
7.10.1;I. INTRODUCTION;126
7.10.2;II. ELIMINATION OF NON-UNIQUENESS;127
7.10.3;III. USE OF ESTIMABLE FUNCTIONS OF PARAMETERS;131
7.10.4;REFERENCES;132
8;Section 2: Automatic Classification and Related Techniques;134
8.1;CHAPTER 11.
SOME RECENT ADVANCES IN CLUSTERING;136
8.1.1;I. INTRODUCTION;136
8.1.2;II. CLASSIFICATION SPACE AND REPRESENTATION SPACE;138
8.1.3;III. LEARNING HIERARCHICAL CLUSTERING FROM EXAMPLES;144
8.1.4;IV. NEW KIND.S OF GRAPHICAL REPRESENTATION IN CLUSTERING;147
8.1.5;REFERENCES;152
8.2;CHAPTER 12.
SIMULTANEOUS CLUSTERING OF CASES AND VARIANCES;154
8.2.1;I. INTRODUCTION;154
8.2.2;II. MODELS AND THEIR FITTING;155
8.2.3;III. STRATEGIES OF DATA ANALYSIS;162
8.2.4;REFERENCES;165
8.3;CHAPTER 13.
TECHNIQUES OF APPROXIMATION FOR BUILDING TWO TREE STRUCTURES;168
8.3.1;I.- INTRODUCTION;168
8.3.2;II.- CONTINUOUS APPROACH OF AN ULTRAMETRIC;169
8.3.3;III.- CONTINUOUS APPROACH OF AN ADDITIVE TREE METRIC;176
8.3.4;IV.- AN EXAMPLE IN ECOLOGY;178
8.3.5;V.- CONCLUSION;186
8.3.6;REFERENCES;186
8.4;CHAPTER 14.
A HIERARCHICAL CLUSTERING METHOD FOR DISSIMILARITY MATRICES WITH INDIVIDUAL DIFFERENCES;188
8.4.1;I. INTRODUCTION;188
8.4.2;II. MODEL;189
8.4.3;III. METHOD;190
8.4.4;IV. ALGORITHM;192
8.4.5;V. EXAMPLE;193
8.4.6;REFERENCES;195
8.5;CHAPTER 15.
APPLICATIONS OF MULTIVARIATE AND CLADISTIC DATA ANALYSES FOR THE CLASSIFICATION OF DERMAPTEROUS INSECTS;196
8.5.1;I. INTRODUCTION;196
8.5.2;II. DERMAPTERAN FOSSILS;196
8.5.3;III. DERMAPTERAN BIOGEOGRAPHY;197
8.5.4;IV. DERMAPTERAN MULTIVARIATE MORPHOMETRICS;197
8.5.5;V. DERMAPTERAN PHYLOGENETIC AND CLADISTIC INFORMATION;200
8.5.6;VI. DERMAPTERAN PHYSICAL TAXONOMY AND FUTURE CLASSIFICATION;201
8.5.7;REFERENCES;201
8.6;CHAPTER 16.
COMPARING RELATIONAL VARIABLES ACCORDING TO LIKELIHOOD OF THE LINKS CLASSIFICATION METHOD;204
8.6.1;I. INTRODUCTION ; REPRESENTATION OF RELATIONAL VARIABLES;204
8.6.2;II. COMPARING RELATIONAL VARIABLES;208
8.6.3;REFERENCES;215
8.7;CHAPTER 17.
ROLE OF COMPUTER GRAPHICS IN INTERPRETATION OF CLUSTERING RESULTS;218
8.7.1;I. INTRODUCTION;218
8.7.2;II. GRAPHICAL REPRESENTATION IN AUTOMATIC CLASSIFICATION;219
8.7.3;III. CONCEPT OF COMPUTER GRAPHICS SYSTEM IN AUTOMATIC CLASSIFICATION;220
8.7.4;IV. CASE STUDIES;224
8.7.5;V. CONCLUSION;237
8.7.6;ACKNOWLEDGEMENTS;238
8.7.7;REFERENCES;238
8.8;CHAPTER 18.
CLASSIFICATION OF FRUIT AND VEGETABLE VARIETIES BY CHEMICAL ANALYSIS OF FRAGRANCE SUBSTANCES;240
8.8.1;I. INTRODUCTION;240
8.8.2;II. MATERIALS AND METHODS;241
8.8.3;III. EXPERIMENTAL RESULTS;241
8.8.4;IV. DISCUSSION;243
8.8.5;REFERENCES;245
9;Section 3: Scaling Method and Correspondence Analysis from the Viewpoint of Practical Approach;246
9.1;CHAPTER 19.
ASSESSING THE NUMBER OF AXES THAT SHOULD BE CONSIDERED IN CORRESPONDENCE ANALYSIS;248
9.1.1;I. INTRODUCTION;248
9.1.2;II. CORRESPONDENCE ANALYSIS AS AN APPROXIMATION OF THE DATA MATRIX;249
9.1.3;III. DETERMINING THE NUMBER OF AXES TO BE RETAINED BY CROSS VALIDATION;253
9.1.4;IV. TESTS ON THE SUM OF THE NON-RETAINED LATENT ROOTS;254
9.1.5;REFERENCES;256
9.2;CHAPTER 20.
SINGULAR VALUE DECOMPOSITION OF MULTIARRAY DATA AND ITS APPLICATIONS;258
9.2.1;I. INTRODUCTION;258
9.2.2;II. SINGULAR VALUE DECOMPOSITION;259
9.2.3;III. ALGORITHM AND CRITERIA;264
9.2.4;IV. NESTED CONFIGURATION AND INTERPRETATION;266
9.2.5;V. SOME APPLICATIONS;269
9.2.6;REFERENCES;273
9.3;CHAPTER 21.
PARTIAL CORRESPONDENCE ANALYSIS AND ITS PROPERTIES;276
9.3.1;I. INTRODUCTION;276
9.3.2;II. MATHEMATICAL PREPARATIONS;277
9.3.3;III. PARTIAL CORRESPONDENCE ANALYSIS;279
9.3.4;IV. NUMERICAL EXAMPLE OF PARTIAL CORRESPONDENCE ANALYSIS;281
9.3.5;REFERENCES;283
9.4;CHAPTER 22.
CORRELATION ANALYSIS OF N-WAY QUALITATIVE DATA AND ITS APPLICATIONS;284
9.4.1;I. INTRODUCTION;284
9.4.2;II. OPTIMAL SCORING METHOD MAXIMIZING CANONICAL CORRELATION COEFFICIENT;285
9.4.3;III. OPTIMAL SCORING METHOD FOR THREE-WAY QUALITATIVE DATA;287
9.4.4;IV. EXAMPLES FOR THREE-WAY QUALITATIVE DATA;290
9.4.5;REFERENCES;296
9.5;CHAPTER 23.
OUTLIERS AND INFLUENTIAL OBSERVATIONS IN QUANTIFICATION THEORY;298
9.5.1;I. INTRODUCTION;298
9.5.2;II. QUANTIFICATION I;299
9.5.3;III. QUANTIFICATION II;303
9.5.4;IV. NUMERICAL INVESTIGATION;305
9.5.5;V. DISCUSSION;308
9.5.6;REFERENCES;310
9.6;CHAPTER 24. CONVERSATIONAL DATA ANALYSIS SYSTEM;312
9.6.1;I. INTRODUCTION;312
9.6.2;II. PRELIMINARY ANALYSIS OF MULTIVARIATE DATA;313
9.6.3;III. CONVERSATIONAL SELECTION OF VARIABLES AND ITEMS;316
9.6.4;IV. REGRESSION DIAGNOSIS AND ANALYSIS OF RESIDUAL;319
9.6.5;V. POOLING OF CATEGORIES;320
9.6.6;VI. CONVERSATIONAL PROCESSING AND BATCH PROCESSING;321
9.7;CHAPTER 25.
ANALYSIS AND COMPARISON OF DIFFERENT TABLES;324
9.7.1;I. INTRODUCTION;324
9.7.2;II. MULTIPLE FACTOR ANALYSIS;325
9.7.3;III. ANALYSIS OF FREQUENCY TABLES;335
9.7.4;CONCLUSION;337
9.7.5;REFERENCE;338
9.8;CHAPTER 26.
USE OF TABULATED DATA IN DATA ANALYSIS;340
9.8.1;I. INTRODUCTION;340
9.8.2;II. CROSS-TABULATION AS A PROCEDURE OF DATA ANALYSIS;341
9.8.3;III. DESIGNING ANALYSIS;345
10;Section 4: Applications: Extraction and Interpretation of Information in Multidimensional Data;346
10.1;CHAPTER 27.
MULTIVARIATE DESCRIPTIVE TECHNIQUES APPLIED TO THE PROCESSING OF LONGITUDINAL SAMPLE SURVEY DATA;348
10.1.1;I. THE INITIAL PROBLEMS;348
10.1.2;II. PRESENTATION OF THE DATA SET;349
10.1.3;III. PRESENTATION OF THE METHODS;349
10.1.4;IV. THE FUNDAMENTAL NOTION OF ACTIVE VARIABLE (AV);349
10.1.5;V. THE "SWARM" OF AV AND THE "GRID" OF SV CATEGORIES;351
10.1.6;VI. CHANGES IN THE PATTERNINGS OF OPINIONS;356
10.1.7;VII. REMARKS : Existence and autonomy of structures;358
10.1.8;REFERENCES;359
10.2;CHAPTER 28.
MULTIDIMENSIONAL ANALYSIS OF OPINION SURVEY DATA;360
10.2.1;I. FORWORD;360
10.2.2;II. METHODS OF ANALYSIS;362
10.2.3;III. Application Examples;369
10.2.4;IV. CONCLUSION;388
10.2.5;REFERENCES;388
10.3;CHAPTER 29. DATA ANALYTIC APPROACHES TO HUMAN BEHAVIORAL RELATIONSHIPS IN A SURVEY OF ACCIDENTS;390
10.3.1;I. INTRODUCTION;390
10.3.2;II. METHOD;391
10.3.3;III. RESULTS AND DISCUSSION;391
10.3.4;ACKNOWLEDGEMENTS;397
10.3.5;REFERENCE;397
10.4;CHAPTER 30.
ABOUT THE NUMERICAL AND STATISTICAL STABILITY OF FORECASTING ALGORITHMS;398
10.4.1;I. THE STATISTICAL MODEL OF EVENTS FORECAST;398
10.4.2;II. ATTENUATION COEFFICIENT OF THE MAHALANOBIS DISTANCE;401
10.4.3;III. THE DISTRIBUTION LAW OF THE MAHALANOBIS DISTANCE;404
10.4.4;IV. NUMERICAL STABILITY OF THE FORECASTING MODEL;406
10.4.5;V. STATISTICAL STABILITY OF THE FORECASTING MODEL;408
10.4.6;VI. THE NATURAL LINK TO OVEREVALUATE THE NUMBER OF PREDICTORS;414
10.4.7;REFERENCES;415
10.5;CHAPTER 31.
THE METHOD OF PATTERN CLASSIFICATION AND ITS APPLICATION TO PROGNOSIS OF DISEASE;418
10.5.1;I. INTRODUCTION;418
10.5.2;II. QUANTIFICATION THEORY TYPE III AND ONE DIMENSIONAL SCALING STRUCTURE;418
10.5.3;III. APPLICATION;425
10.5.4;IV. CONCLUSION;426
10.5.5;REFERENCES;428
10.6;CHAPTER 32.
A NON PARAMETRIC DISCRIMINANT ANALYSIS BASED ON THE CONSTRUCTION OF A BINARY DECISION TREE;430
10.6.1;I. INTRODUCTION;430
10.6.2;II. ILLUSTRATIVE BINARY DECISION TREE;431
10.6.3;III. CONSTRUCTION OF A BINARY DECISION TREE;431
10.6.4;IV. PRUNING PROCESS;435
10.6.5;V. CHOOSING THE RIGHT-SIZED TREE : A PROBLEM OF DETERMINATION OF THE MOST RELIABLE ESTIMATED TRUE ERROR RATE;437
10.6.6;VI. CONCLUSION;441
10.6.7;REFERENCES;441
10.7;CHAPTER 33.
EXPERIMENTAL COMPARISON BETWEEN THE OPTIMAL DISCRIMINATE PLANE BASED ON SAMPLES AND GENERAL DISCRIMINANT ANALYSIS;442
10.7.1;I. INTRODUCTION;442
10.7.2;II. THEORETICAL CONSIDARATION;443
10.7.3;III. EXPERIMENT;448
10.7.4;IV. CONCLUSION;452
10.7.5;REFERENCES;452
10.8;CHAPTER 34.
A METHOD OF DISCRIMINATION FOR ELECTRICAL BIOSIGNAL;454
10.8.1;I. INTRODUCTION;454
10.8.2;II. SPIKE DETECTION;454
10.8.3;III. FETAL QRS DETECTION;457
10.8.4;IV. DISCUSSION;459
10.8.5;V. CONCLUSION;461
10.8.6;ACKNOWLEDGEMENT;461
10.8.7;REFERENCES;461
10.9;CHAPTER 35.
CLASSIFICATION OF BEHAVIORAL STATES OF THE MOUSE;462
10.9.1;I. MICRO STAGE AGGREGATION PROCEDURE;462
10.9.2;II. MACRO ANALYSIS;466
10.9.3;REFERENCE;469