The Construction of Financial Distress Prediction Model Based on Hybrid Strategy

Aiming at the unbalanced datasets in financial distress prediction research,this paper proposes a model based on hybrid strategy.The main idea is to provide effective modeling program with a combination of preprocessing data-level and ensemble-learning algorithm-level.At first,by re-sampling the financial distress samples using SMOTE algorithm,a serious loss of information caused by selecting the matched samples in the traditional study of financial distress prediction can be avoided.Then by training the sample datasets using Bagging-C4.5 algorithm,the prediction accuracy of financial distress samples can be greatly enhanced.Experimental results show that this model is valid for financial distress prediction and also of practical significance.