Dynamic Fuzzy Pattern Recognition With Applications To Finance And Engineering (international Series In Intelligent Technologies)

__Dynamic Fuzzy Pattern Recognition with Applications to Finance and____Engineering__ focuses on fuzzy clustering methods which have proven to be very powerful in pattern recognition and considers the entire process of dynamic pattern recognition. This book sets a general framework for Dynamic Pattern Recognition, describing in detail the monitoring process using fuzzy tools and the adaptation process in which the classifiers have to be adapted, using the observations of the dynamic process. It then focuses on the problem of a changing cluster structure (new clusters, merging of clusters, splitting of clusters and the detection of gradual changes in the cluster structure). Finally, the book integrates these parts into a complete algorithm for dynamic fuzzy classifier design and classification. Contents......Page 3 Figures......Page 7 Tables......Page 12 Foreword......Page 14 1 Introduction......Page 16 Goals & Tasks of the Book......Page 17 Structure of the Book......Page 19 Knowledge Discovery Process......Page 21 Problem of Pattern Recognition......Page 25 Problem of DPR......Page 34 Monitoring Process......Page 51 Adaptation Process......Page 71 4 Dynamic Fuzzy Classifier Design with Point-Prototype based Clustering Algorithms......Page 92 Formulation of Dynamic Clustering......Page 93 Requirements for Clustering Algorithm used for Dynamic Clustering & Classification......Page 99 Detection of New Clusters......Page 101 Merging of Clusters......Page 113 Splitting of Clusters......Page 131 Detection of Gradual Changes in Cluster Structure......Page 145 Adaptation Procedure......Page 146 Updating Template Set of Objects......Page 149 Cluster Validity Measures for Dynamic Classifiers......Page 156 Summary of Algorithm for Dynamic Fuzzy Classifier Design & Classification......Page 164 5 Similarity Concepts for Dynamic Objects in Pattern Recognition......Page 167 Extraction of Characteristic Values from Trajectories......Page 169 Similarity Notion for Trajectories......Page 172 Extension of Fuzzy Pattern Recognition Methods by Similarity Measures for Trajectories......Page 209 Bank Customer Segmentation based on Customer Behaviour......Page 211 Computer Network Optimisation based on Dynamic Network Load Classification......Page 245 7 Conclusions......Page 276 Refs......Page 279 Unsupervised Optimal Fuzzy Clustering Algorithm of Gath & Geva......Page 288 Description of implemented Software......Page 291 Index......Page 294

Dynamic Fuzzy Pattern Recognition with Applications to Finance and Engineering focuses on fuzzy clustering methods which have proven to be very powerful in pattern recognition and considers the entire process of dynamic pattern recognition. This book sets a general framework for Dynamic Pattern Recognition, describing in detail the monitoring process using fuzzy tools and the adaptation process in which the classifiers have to be adapted, using the observations of the dynamic process. It then focuses on the problem of a changing cluster structure (new clusters, merging of clusters, splitting of clusters and the detection of gradual changes in the cluster structure). Finally, the book integrates these parts into a complete algorithm for dynamic fuzzy classifier design and classification.