Molecular recognition analysis not only provides basic information of compositions and structures of biomacromolecules for molecular design of drugs, but also gives models for understanding the interactive modes and sites between drug molecules and biomacromolecules.

In this model, four multiple discriminant analysis (MDA) criterion functions are defined and used to achieve global optimization in finding the best segmentation by means of the smallest within-segment distance, the largest between-segment distance and segment length.

In our research we employed CART(Classify and Regression Trees) and MDA(Multiple Discriminant Analysis) to establish two models with annual financial statements of China\'s 196 listed companies in 2006,and then we evaluated the model\'s accuracy with the data from 1624 listed companies in 2007.

Based on the real data sets which involves different periods and different sizes of cities from national commercial banks, this paper applied the methodologies of Multivariate Discrimination Analysis, Logistic Regression, Decision Tree and Neural Network to accomplishing the detailed empirical research.