QTOP-K: A novel Algorithm for mining high quality pattern-based clusters in GST Microarray Data

Pattern-based clustering is widely applied in bioinformatics and biomedical Recently, mining high quality pattern-based clusters has become an important research direction. However, the existing methods were neither efficient in large data set nor precise at measuring the quality of clusters. These problems have greatly limited the methods´ application in large data set. This paper proposes a new algorithm, which can provide a more accurate measurement for the quality of clusters and sharply cut down the time for mining high quality patterned-based clusters compared with today´s methods. Experiments are held on real data set and synthetic data set and the test result suggests that Qtop-k has made notable progress in the aforementioned problems