Synthetic oversampling for credit card default prediction

This dataset contains more than 17000 data of credit card holder with 20 predictor variables and 1 binary target variable. The corresponding R code for comparing several proposed (density-based) and existing synthetic oversampling methods (SMOTE-based) is also provided.本数据集包含超过17000条信用卡持卡人数据，共涵盖20个预测变量（predictor variables）与1个二分类目标变量（binary target variable）。本数据集还附带了用于对比多种本文提出的基于密度的合成过采样方法与现有基于SMOTE（Synthetic Minority Oversampling Technique）的合成过采样方法的R代码。