Back to search

Publikationsserver der RWTH Aachen University

Dynamic fuzzy pattern recognition

Abstract

dc:description

Dynamic pattern recognition is concerned with the recognition of clusters of dynamic objects, i.e. recognition of typical states in the dynamic behaviour of a system under consideration. The goal of this thesis is to investigate this new field of dynamic pattern recognition and to develop new methods for clustering and classification in a dynamic environment. The first topic of investigation is the procedure of dynamic classifier design enabling a design of an adaptive classifier that can automatically recognise new cluster structures as time passes. The algorithm proposed in this thesis is introduced in the framework of unsupervised learning and takes advantage of fuzzy set theory which provides a unique mechanism for gradual assignment of objects to clusters and allows to detect a gradual temporal transition of objects between clusters. Another topic addressed in this book is the definition of similarity measures for trajectories of dynamic objects. A number of similarity measures for trajectories are proposed which take into consideration either the pointwise closeness of trajectories in the feature space or the best match of trajectories with respect to their shape. In order to demonstrate the practical relevance of the algorithm developed here, it was applied to two different problems. The first economic problem is concerned with bank customer segmentation based on the analysis of the customers' behaviour. The second technical problem is related to the recognition of typical states in computer network load. The target groups of this book are scientists and practitioners of the very new and growing area of dynamic data analysis, graduate and PhD students of operations research, computer science, and engineering.

Degree

thesis:*
Grantor dc:publisher
Publikationsserver der RWTH Aachen University
Year dc:date
2001

Author and committee

dc:creator, dc:contributor.*
Author dc:creator
  • Angstenberger, Larisa
Contributors dc:contributor
  • Zimmermann, Hans-Jürgen

Subjects

dc:subject × 8

Rights

dc:rights
Statement dc:rights
  • info:eu-repo/semantics/openAccess
Language dc:language
eng

Identifiers

dc:identifier.*
OAI identifier oai:identifier
oai:publications.rwth-aachen.de:59797

Chain of custody

source
Harvested from
RWTH Aachen University
Base URL
publications.rwth-aachen.de/oai2d
Last updated
2026-07-30
Source record
OAI-PMH GetRecord
citation

Angstenberger, Larisa. Dynamic fuzzy pattern recognition. Publikationsserver der RWTH Aachen University, 2001. https://publications.rwth-aachen.de/record/59797