Based on the existing researches,we use dual membership fuzzy support vector machine(SVM) model to analyze listed companies financial distress.In this model,each sample belongs to two classes according to its dual membership.To solve the class imbalance problem in financial distress,we propose a new SVM model called CI-FSVM.Empirical results show that the discrimination accuracy of CI-FSVM is significantly better than other existing SVM models and the new model has great application value.