We propose a novel scheme for designing fuzzy rule based classifier. An SOFM based method is used for generating a set of prototypes which is used to generate a set of fuzzy rules. Each rule represents a region in the feature space that we call the context of the rule. The rules are tuned with respect to their context. We justified that the reasoning scheme may be different in different context leading to context sensitive inferencing. To realize context sensitive inferencing we used a softmin operator with a tunable parameter. The proposed scheme is tested on several multispectral satellite image data sets and the performance is found to be much better than the results reported in the literature.
updated: Mon Nov 23 2009 14:26:00 GMT+0000 (UTC)
published: Mon Nov 23 2009 14:26:00 GMT+0000 (UTC)