cancer diagnosis and treatment involves discovering and classifying cancer types. Most of the previous works involve the single clustering algorithms. In Golub’swork [3], the self-organizing feature map and neighbourhood analysis were adopted to discover two types of human acute leukemia, which are acute myeloid leukemia (AML) and acute lymphoblastic leukemia (ALL). In Wigle’s work [4], clustering approaches and statistical analysis were adopted to identify non-small cell lung cancer (NSCLC) from the normal cases. These works include certain limitations such as lack of robustness, stability and accuracy. But in our paper, we have adopted the concept called Hierarchical clustering which is one of the multiple clustering algorithms. This is the powerful method for improving both the robustness as well as the stability of unsupervised classification solutions.