Credit Scoring Based on Kernel Matching Pursuit

Credit risk is paid more and more attention by financial institutions, and credit scoring has become an active research topic. This paper proposes a new credit scoring method based on kernel matching pursuit (KMP). KMP appends sequentially basic functions from a kernel-based dictionary to an initial empty basis using a greedy optimization algorithm, to approximate a given function, and obtain the final solution with a linear combination of chosen functions. An outstanding advantage of KMP in solving classification problems is the sparsity of its solution. Experiments based on two real data sets from UCI repository show the effectiveness and sparsity of KMP in building credit scoring model.