Diagnosis of Breast Cancer by Using Competitive Learning with Modified Confidential Voting
Itani NAGAYAMA, Morikazu Nakamura, Tsuyoshi Yamashiro · IEEJ Transactions on Industry Applications · 2001
This paper describes a new approach for competitive learning network and its application to diagnosis of breast cancer in medical pattern recognition. Conventional competitive learning such as RCE network proposed by Cooper et. al. has a good performance caused by the faster learnability than popular steepest descent methods. However, the RCE network has a shortcoming with its decision making process. Thus, we introduce the confidential voting process in order to avoid the ambiguous decision making in the RCE network. Generally speaking, despite of great deal of public awareness, breast cancer continues to be one of the most common cancer diseases. Therefore, if we can find an efficient method to diagnose the cancer with computer technology, the more efficient early detection can be performed. We first discuss some essential issues to be considered in medical aspects of cancer diagnosis. Experimental results by using the proposed competitive network with confidential voting are also described. The good performance of the new competitive learning approach is shown.