Credit scoring with class imbalance data: An out-of-sample and out-of-time perspective.

Data for predicting default risk in performance windows ranging from 12 to 60 months. For example, FM12 folder contains data on whether mortgage loan profiles with a credit score of less than 730 defaulted within 12 months of the loan origination date. A training sample is provided for each outcome window, as well as out-of-sample (OOS) and out-of-time (OOT) testing samples. OOS is generated using the same timeline as the training sample, whereas OOT is created from a future timeline. The samples are all generated using common features.

本数据集用于预测12至60个月贷款表现窗口期内的违约风险。例如，FM12文件夹收录了信用评分低于730的抵押贷款相关数据，记录了此类贷款在贷款发放之日起12个月内是否发生违约。针对每一类结果窗口期，均提供了训练样本，以及样本外（out-of-sample, OOS）测试样本与跨时间（out-of-time, OOT）测试样本。其中样本外测试样本与训练样本采用一致的时间线生成，而跨时间测试样本则取自未来的时间区间。所有样本均通过通用特征构建生成。