Non-standard parameter adaptation for exploratory data analysis

Abstract

This book presents the fruits of several years of research into non-standard methods of adaptation for exploratory data analysis. This research resulted in the award of PhD to two of the authors, Dr Wesam Barbakh and Dr Ying Wu, at the University of the West of Scotland in early 2008. Their respective areas of research are Dr Barbakh is mainly interested in clustering and topology preserving mappings. In particular his work concentrates on formulating diverse performance functions which incorporate local and global information in order to ensure that prototypes attain a more globally optimal convergence. These non-standard performance functions have translated into non-standard adaptation algorithms. Dr Wu is the most directly intereted in non-standard adaptation per se. He has extensive experience of non-standard exploratory data analysis using such methods as reinforcement learning and the cross entropy …