Building Default Predicting Model on Firm´s Short-term Loan Data with Artificial Neural Network - Considering Qualitative Indexes and Misclassification Costs

To date, using model to predict whether firm´s default is still a problem. It presents: a. most model using pairwise pattern; b. lack of qualitative indexes that affect firm´s default; c. asymmetric between normal firm´s misclassification costs and default firm´s. So, introducing qualitative indexes, using all samples and considering misclassification costs, this paper builds an artificial neural network model on short-term-loan data. Though training, validating and testing, it´s performance is good.