Article Details

Abstract

The rapid development of the E-Commerce industry makes customer churn prediction increasingly important. This research focuses on using Machine Learning technology to address the customer churn early warning problem in the E-Commerce domain. Based on the Alibaba Cloud Tianchi dataset, this paper uses the Random Forest algorithm to construct a prediction model and compares it with other classical Machine Learning algorithms such as Naive Bayes, K-Nearest Neighbors, and Decision Tree, verifying the application value of the Random Forest model in this field. The experimental results show that the Random Forest model achieves the best performance on the customer churn prediction task, significantly outperforming other comparison models. In the balanced dataset experiment, its AUC value is also higher than those of the Naive Bayes algorithm, K-Nearest Neighbors classification algorithm, and Decision Tree model, fully demonstrating the excellent discriminative ability of the Random Forest model in distinguishing between churn and non-churn customers. This proves the effectiveness of the Random Forest model for E-Commerce enterprises to build an efficient and reliable customer churn early warning system, providing a methodological reference for the construction of early warning systems in E-Commerce enterprises.

Keywords

customer churn prediction; E-Commerce; Random Forest; Machine Learning

References

1. [1] Wu Junzhuo, Wang Mingwen, Yu Zhiyuan, etc Customer churn prediction model based on graph Transformer [J]. Journal of Jiangxi Normal University (Natural Science Edition), 2025, 49 (4): 378-391432
2. [2] Jia Chengrui Defects of Random Forest Model in Customer Loss Prediction [J]. National Circulation Economy, 2021 (4): 9-11
3. [3] YULIANTI Y，SAIFUDIN A.Sequential Feature Selection in Customer Churn Prediction Based on Naive Bayes [J/OL].IOP Conference Series: Materials Science and Engineering，2020，879（1）：012090[2025-10-14].https://iopscience.iop.org/article/10.1088/1757-899X/879/1/012090.
4. [4] ETAIWI W，BILTAWI M，NAYMAT G. Evaluation of classification algorithms for banking customer's behavior under Apache Spark Data Processing System [J].Procedia Computer Science，2017，113：559-564.
5. [5] MANZOOR A，QURESHI M A，KIDNEY E， et al. A Review on Machine Learning Methods for Customer Churn Prediction and Recommendations for Business Practitioners [J].IEEE Access，2024，12：70434-70463.
6. [6] Shi Wenbing, Liu Liyuan, Liu Yaping Research on Ecological Environment Quality of Saihanba Mechanical Forest Farm Based on Random Forest Classification [J]. Environmental Science and Management, 2025, 50 (6): 189-194
7. [7] DU M N，LIU N H，HU X. Techniques for interpretable machine learning [J/OL].arXiv:1808.00033 [cs.LG].[2025-10-05].https://doi.org/10.48550/arXiv.1808.00033.
8. [8] JAGADEESH B，SASHIREKHA K. Enhancing Accuracy of Customer Churn Prediction in e-commerce Using Random Forest Algorithm over K-nearest Neighbors Algorithm [C]//International Conference on Newer Engineering Concepts and Technology: ICONNECT-2024.Trichy：AIP，2024：3270.
9. [9] Qiao Jian, Zhu Jiahui, Yan Kanghuan A Telecom Customer churn Prediction Model Based on Improved Random Forest CART Feature Selection Algorithm [J]. Telecommunication Engineering Technology and Standardization, 2022, 35 (3): 78-82
10. [10] Yang Lin, Bai Zhao, Kou Yonggang Random Forest Algorithm Based on RFM Model for Customer Loss Analysis in Civil Aviation [J]. Computer and Modernization, 2021 (1): 100-104
11. [11] Zhang Miao, Gu Haiyan Prediction of national GDP using a random forest model based on optimized decision tree parameters [J]. China Forestry Economy, 2025 (4): 39-50
12. [12] Teacher Dong, Huang Philosophy Analysis of Random Forest Theory [J]. Ensemble Techniques, 2013, 2 (1): 1-7

About the authors

Zhu Zibaijia (2004-), male, Han ethnicity, from Liuyang, Hunan, is currently pursuing a bachelor's degree with a research focus on economics and management.