Using a modified counter-propagation algorithm to classify conjoint data

Authors:
Hans Pierrot;Tim Hendtlass
Affiliations:
Centre for Intelligent Systems and Complex Processes, Swinburne University of Technology, Hawthorn, Australia;Centre for Intelligent Systems and Complex Processes, Swinburne University of Technology, Hawthorn, Australia
Venue:
IEA/AIE'2003 Proceedings of the 16th international conference on Developments in applied artificial intelligence
Year:
2003

Citing 1
Cited 0

Fast learning in networks of locally-tuned processing units

Neural Computation

Quantified Score

Hi-index	0.00

Visualization

Abstract

Conjoint data is data in which the classes abut but do not overlap. It is difficult to determine the boundary between the classes as there are no inherent clusters in conjoint data and as a result traditional classification methods, such as counter propagation networks, may under perform. This paper describes a modified counter propagation network that is able to refine the boundary definition and so perform better when classifying conjoint data. The efficiency with which it uses the network resources suggests that it is worthy of consideration for classifying all kinds of data.