To Reduce Gross NPA and Classify Defaulters Using Shannon Entropy

Non-performing asset (NPA) has been in a serious attention by banks over the past few years. NPA causes huge loss to banks; hence, it becomes an extremely critical step deciding which loans have capabilities become NPA, and thereby grant ones reject. In this paper, focuses on exact crux of matter, algorithm is proposed designed handle financial data very meticulously predict with high accuracy whether particular loan would be classified as future or not. Instead conventional, less accurate classifiers used decide can turn unique modeling entropy-based classifier model built using logic Shannon entropy. The categorizes points two categories, ‘accepted’ ‘rejected’. use local entropy global also exerted help determine output. then compared existing NPAs, giving idea about performance.