贷款违约预测

该赛题数据来自某信贷平台的贷款记录，总数据量超过120万条，从中抽取80万条作为训练集，20万条作为测试集A，20万条作为测试集B。每条数据包含47列变量信息，其中15列为匿名变量，具体如下表，同时会对employmentTitle、purpose、postCode和title等信息进行脱敏。

This competition dataset is sourced from loan records of a credit platform. The total number of records exceeds 1.2 million, with 800,000 selected as the training set, 200,000 as test set A, and the remaining 200,000 as test set B. Each record contains 47 columns of variable information, 15 of which are anonymous variables as detailed in the following table. Furthermore, information fields including employmentTitle, purpose, postCode and title have been desensitized.

搜集汇总

背景与挑战

背景概述

该数据集是一个用于金融风控贷款违约预测的分类任务数据集，数据来源于某信贷平台，总规模超过120万条记录，包含80万条训练数据和40万条测试数据。每条数据有47个特征变量，其中15个为匿名变量，并对就业标题、贷款目的等敏感信息进行了脱敏处理，适用于构建违约预测模型。