On Finding the Best Parameters of Fuzzy k-Means for Clustering Microarray Data
Wei Yang, Luis Rueda and Alioune Ngom
Clustering algorithms, such as hierarchical clustering, k-means, and fuzzy k-means, have become important tools for gene expression analysis of microarray data. However, the need of prior knowledge of the number of clusters, k, and the fuzziness parameter, b, limits the usage of fuzzy clustering. In this paper, we use simulated annealing (SA) and fuzzy k-means clustering to determine the optimal parameters, namely the number of clusters, k, and the fuzziness parameter, b. To improve SA, two methods, searching with Tabu List and shrinking the scope of randomization, are applied. Our results show that a nearly-optimal pair of k and b (a near optimal value of k) can be obtained without exploring the entire search space.