# DSC Bi-Weekly Challenge #2 – Customer Retention — Data

来源：https://www.kaggle.com/competitions/dsc-contest-02/data

采集时间：2026-10-20T00:45:27Z

The datasets provided for this competition (both training and testing) are designed to closely resemble real-world banking customer data. While the data reflects realistic customer behavior patterns, the feature distributions may differ slightly from commonly available public datasets, encouraging participants to focus on robust modeling and generalization rather than memorization.

Participants are welcome to explore publicly available bank churn datasets for additional context, comparison, and experimentation. Such datasets may help in understanding feature behavior and improving model performance, provided that competition rules are followed.

## Files

*   **train.csv** - the training dataset; `Exited` is the binary target
*   **test.csv** - the test dataset; your objective is to predict the probability of `Exited`
*   **sample_submission.csv** - a sample submission file in the correct format