Profitability Prediction Model Based on Support Vector Machines

Classification techniques which hold a very important place in business and economic research have drawn many research interests in previous literatures. Recent studies have shown that machine learning techniques can achieve better performance. However, the application of support vector machines (SVM) for profitability prediction has not been widely explored. This study investigates the efficacy of applying SVM in profitability prediction problem and attempts to suggest a new model with better explanatory power and stability. The experimental results demonstrate that the proposed classifier of SVM approach to the problem of profitability prediction of companies upon the basis of a set of financial ratios.