We explore in this paper a novel clustering algorithm, named CORE (standing for CORrelated-Force Ensemble), for categorical data. In general, it is more difficult to perform clustering on categorical data than on numerical data due to the absence of the ordered property in the former. Though several clustering algorithms which concentrate on categorical date were proposed, acquiring the desirable quality remains a challenging issue. Note that there is significance hidden in the correlation between attribute values that can be explored to aid clustering, especially extracting clusters in the high dimensional data. Therefore by employing the concept of correlated-force ensemble, clusters which consist of the highly correlated set of nominal attribute values, can be acquired by the proposed algorithm, CORE. As validated by variant real datasets, it is shown in our experimental results that algorithm CORE significantly outperforms the prior works.
|出版狀態||Published - 2004 1月 1|
|事件||Proceedings of the Fourth SIAM International Conference on Data Mining - Lake Buena Vista, FL, United States|
持續時間: 2004 4月 22 → 2004 4月 24
|Other||Proceedings of the Fourth SIAM International Conference on Data Mining|
|城市||Lake Buena Vista, FL|
|期間||04-04-22 → 04-04-24|
All Science Journal Classification (ASJC) codes