Classification of post operative breast cancer patient information using complex valued neural classifiers

Classification of Haberman's Survival information is useful to find out the patients survival probability after a breast cancer surgery. Dataset has been collected from a standard benchmark UCI machine learning repository. A study at the hospital named University of Chicago's Billings was conducted between the year 1958 and 1970 to identify the cancer patients who had undergone surgery for breast cancer and survived. The data obtained are classified using a fully complex valued classifier in this paper. Classifying patient's survival after five years and patients death within five years is a challenging prognosis problem. The effectiveness of the classification achieved can be used by the clinicians for the treatment of patients in the hospitals. For achieving better discrimination, the proposed method uses a fully complex valued fast learning classifier with Gd activation function in the hidden layer. Comparing the classification efficiency of FC-FLC with other networks available in the literature, FC-FLC provides a better classification performance than the SRAN, MCFIS and ELM classifier.