Clustering method based on λ-Warshall algorithm
Feng Yan · Jisuanji gongcheng yu sheji · 2008
Clustering method based the improved Warshall algorithm is proposed by using the similarity of data.The method on the basis of the traditional Warshall algorithm forms the transitive closure of fuzzy similar relation by using the clustering parameter.The transitive closure of fuzzy similar relation is the equivalence closure owing to the reflexive and symmetric properties of fuzzy similar re-lation.The data with the equivalence relation are divided into the same clusters.The experimental results show that the clustering al-gorithm achieves the same cluster results comparing with the traditional k-means clustering algorithm.