基于随机森林的个人信用评估模型研究及实证分析

FeaturesofRandomForestsItisunexcelledinaccuracyamongcurrentalgorithms.Itrunsefficientlyonlargedatabases.Itcanhandlethousandsofinputvariableswithoutvariabledeletion.Itgivesestimatesofwhatvariablesareimportantintheclassification.Itgeneratesaninternalunbiasedestimateofthegeneralizationer