决策树在贷款客户信用评估中的应用

针对如何在贷款过程中尽可能地降低风险,提出利用决策树理论对客户的基本情况进行分析,建立银行贷款客户信用评估决策树模型。为防止过拟合问题的产生,对最初生成的决策树又利用基于规则的方法进行裁剪修正,使得决策模型不只在训练集上的信用度决策获正确率较高,在相互独立的不同测试集和验证集上都取得令人满意的效果。

Application of Decision Tree in the Loan Evaluation of Customer Credit

For reducing the risk in the loan process,proposes using decision tree theory to analyze the basic information of the customers and constructs the customer credit evaluation decision tree model.To prevent the over-fitting problems,uses the method based on rule to cut and amend the decision tree first generated,to make the decision model is not only in the training set getting the high accuracy of credit decisions but in the independent test set and validation sets can get satisfactory results.