层次分析法和支持向量机在个人信用评估中的应用

Application of Analytic Hierarchy Process and Support Vector Machine in Personal Credit Evaluation

信用贷款在国内正蓬勃开展,但由于目前我国的征信系统还不完善,银行因信息不对称不断提高个人信贷的标准,步骤也很繁琐,从而导致信用贷款的受益面不广。针对此现状,本文着力研究个人信用评估指标的定量化,并在评估方法上进行了创新。论文首先选取了合适的评估指标,完成了定性指标的定量化;为刻画评估指标作用的主次性,利用层次分析法计算出指标权重,依权重来处理数据;最后用支持向量机对数据进行分类。实验结果表明,采用层次分析法进行数据预处理大大提高了个人信用评估预测的准确度。

Credit loan is quite popular domestically nowadays. However,since credit reference system in China is still not that mature. Banks begin to raise the standards of personal credit loans. The steps are also very tedious,which causes the narrowness of benefits. Based on the situation, the quantification of indexes of personal credit loans is studied. It also innovates on the evaluation methods in personal credit loans. Firstly,appropriate indexes are chosen and the quantification of the indexes is finished. Secondly,the weight of each index is calcuated via AHP,by which it processes the data. Thirdly,the data is classified by SVM. The results show that using AHP to process the data greatly improves the accuracy of personal credit evaluation.