An Empirical Study of Financial Distress Prediction of Listed Companies Based on Support Vector Machine

Based oil the drawbacks of the traditional prediction model, this paper discusses the role of Support Vector Machines model in the financial distress prediction of Listed Companies. By some empirical comparison and model analysis of Support Vector Machines(SVM for short), Multiple Discriminant Analysis(MDA for short), Multiple Logistic Analysis(MLA for short) and BP Artificial Neural Networt (BP-ANN for short), we conclude that the average prediction error rate calculated from the 20 groups of testing samples is the lowest by SVM, and is notably advantageous to MDA. MLA and BP-ANN. we also prove the effectiveness and Superiority of SVM model in financial distress prediction.