Clustering Using Rough - Set Feature Selection

`Feature selection aims to remove features unnecessary to the target concept. Rough-set theory (RST) eliminates unimportant or irrelevant features, thus generating a smaller (than the original) set of attributes with the same, or close to, classificatory power. Clustering, also a form of data grouping, groups a set of data such that intra-cluster similarity is maximized and inter-cluster similarity is minimized. As with classification, there exists a group of attributes or features on the basis of which clustering is carried out; hence RST may be used for clustering. This paper analyses the effects of rough sets on clustering using 10 datasets, each including a decision attribute. This generates a framework for applying rough-sets for clustering purposes. Rough-sets are then used for knowledge discovery in clustering and the conclusion indicated a very significant result that removal of individual numeric attributes has far more effect on clustering accuracy than removal of categorical attributes.