Clustering of Leukemia Patients via Gene Expression Data Analysis

This thesis attempts to cluster some leukemia patients described by gene expression data, and discover the most discriminating a few genes that are responsible for the clustering. A combined approach of Principal Direction Divisive Partitioning and bisect K-means algorithms is applied to the clustering of the selected leukemia dataset, and both unsupervised and supervised methods are considered in order to get the optimal results. As shown by the experimental results and the predefined reference, the combination of PDDP and bisect K-means successfully clusters the leukemia patients, and efficiently discovers some significant genes that can serve as the discriminator of the clustering. The combined approach works well on the automatic clustering of leukemia patients depending merely on the gene expression information, and it has great potential on solving similar problems. The discovered a few genes may provide very important information for the diagnosis of the disease of leukemia