A Clustering Algorithm for Weighted Graph Based on Minimum Cut

Clustering is the unsupervised classification of patterns into groups. It groups a set of data in a way that maximizes the similarity within clusters and minimizes the similarity between two different clusters. In this paper, the Hartuv and Shamirpsilas clustering algorithm for similarity graph is extended to the weighted similarity graph. The algorithm has the advantage of many existing algorithm: low polynomial complexity, the provable properties, and automatically determining the number of clusters in the process of clustering. The algorithm is tested on random graph and the experimental results show that the algorithm performs well.