EVALUATION OF NEURAL NETWORK BASED CLASSIFICATION SYSTEMS FOR CLINICAL CANCER DATA CLASSIFICATION
Khalid Mumtaz, Asan Memorial, K. Duraiswamy · 2008
The two important tasks of Data mining are clustering and classification. Breast cancer represents the second leading cause of cancer deaths in women today. Cancer cells are to be classified as either malignant or benign. This paper aims at the study of suitable machine learning technique for multidimensional, clinical cancer data classification. In this paper, three kinds of neural network based classification systems are evaluated for the proposed cancer data classification problem. The evaluated models are: 1.Adaptive Resonance Theory Based Neural Network (ART), 2.Self Organizing Map Based Neural Network (SOM) and 3.Back Propagation Neural network (BPN).