Transactions on Case-Based Reasoning for Multimedia Data
Volume 1 - Number 1 - October 2008 - Pages 21-36
Case-Based Reasoning Adaptation for High Dimensional Solution Space
Y. Zhang1, P. Louvieris1 and M. Petrou2
1Surrey Defence Technology Centre, University of Surrey, Guildford, UK
2Electrical and Electronic Engineering, Imperial College, London, UK
Case-Based Reasoning (CBR) is a methodology that reuses the solutions of previous similar problems to solve new problems. Adaptation is the most difficult stage in the CBR cycle, especially, when the solution space is multi-dimensional. This paper discusses the adaptation of a high dimensional solution space and proposes a possible approach to it. A Visualisation induced Self Organising Map (ViSOM) is used to map the problem space and solution space first, then a Back Propagation (BP) network is applied to analyse the relations between these two maps. A simple military scenario is used as a case study for evaluation purposes.
Download Paper (377 KB)