Enterprise Bankruptcy Prediction Using Noisy-Tolerant Support Vector Machine

Enterprise bankruptcy forecasting is very important to manage credit risk and a lot of scholars applied themselves to study how to increase the accuracy of bankruptcy forecast which requires a powerful learning machine algorithm capable of good generalization on financial data. Therefore, classification algorithms like support vector machine (SVM) are popular for modeling and predicting corporate distress. However, making inferences and choosing appropriate responses based on incomplete, uncertainty and noisy data is challenging in financial settings particularly in bankruptcy prediction. In this paper, we propose a new approach for enterprise bankruptcy prediction, which uses a novel support vector machine and K-nearest neighbor (KNN-SVM) to remove noisy training examples. The experimental results show that the generalization performance and the accuracy of classification are improved significantly compared to that of the traditional SVM classifier, and adapt to engineering applications.