The Application of an Improved K-Means Clustering Method in Microarray Gene Expressing Data

this paper presents a method which improves the traditional K-means clustering. First, it selects the centers of the K-means clustering method using the idea of the method based on density, then, it puts the microarray profiles into several clusters, draws the curve line of a kind of function which assesses the clustering result under different number of clusters and compares with the traditional method. Finally, it predicts a fit number of clusters using another form of the assessment function and draws the clustering result. The experiment carried out in this paper shows that the improved method surpasses the traditional one.