A Stacking-Based Heterogeneous Ensemble Model for Customer Churn Prediction: Synergistic Integration of LightGBM and AdaBoost
Ziyu Zeng （ Chengdu International Studies University,Chengdu Sichuan 611844, China ）
Maoxin Li （ Chengdu International Studies University,Chengdu Sichuan 611844, China ）
Yuyi Huang （ Chengdu International Studies University,Chengdu Sichuan 611844, China ）
Zhaohong Cao （ Chengdu International Studies University,Chengdu Sichuan 611844, China ）
Cunjue Li （ Chengdu International Studies University,Chengdu Sichuan 611844, China ）
https://doi.org/10.37155/2972-4813-gep0402-1Abstract
Customer churn prediction, typically framed as an imbalanced binary classification problem, poses significant challenges to traditional machine learning models and single ensemble methods, which often suffer from limitations in both predictive accuracy and model interpretability. To address these issues, this paper proposes a heterogeneous ensemble learning framework based on Stacking, which integrates LightGBM and AdaBoost as base learners to leverage their complementary strengths in computational efficiency and classification performance. The proposed model employs a five-fold cross-validation strategy to generate meta-features, thereby enhancing generalization capability. Experimental results demonstrate that the Stacking model achieves an AUC of 0.9132, representing a substantial improvement of 11.45% over standalone LightGBM and 8.69% over AdaBoost. Moreover, the model attains a recall rate of 0.9388, effectively aligning with the business priority of minimizing customer churn through high sensitivity. The innovation of this study lies in three aspects: (1) the design of a heterogeneous ensemble architecture that facilitates performance synergy; (2) the use of cross-validation for robust meta-feature generation; and (3) the incorporation of feature importance analysis to enhance model interpretability. The findings validate the effectiveness of the Stacking ensemble in customer churn prediction and provide both theoretical insights and practical guidance for developing intelligent customer relationship management systems.
Keywords
Customer churn prediction; Stacking ensemble learning; LightGBM; AdaBoost; Imbalanced classificationFull Text
PDFReferences
[2] Wang Jian, Zhang Yu. A Study on Methods for Handling Imbalanced Data in Customer Churn Prediction in the Telecommunications Industry [J]. Journal of Management Engineering, 2019, 33(2): 112-120.
[3] Zhou Zhihua. Research Progress in Semi-Supervised Learning [J]. Journal of Computers, 2019, 42(3): 481-507.
[4] Zhang Changwei. Research and Implementation of a Multi-Method Fusion-Based Model for Telecommunications Customer Churn Prediction [D]. Guangdong: South China University of Technology, 2020.
[5] Xue Bing. A Study on Customer Churn Prediction for Telecommunications Operators Based on Multi-Model Fusion [D]. Dongbei University of Finance and Economics, 2022. DOI:10.27006/d.cnki.gdbcu.2022.000579.
[6] Geng Yu. Customer Churn Prediction Based on BO-Stacking Ensemble Learning [J]. Science and Industry, 2025, 25(13): 241-245.
[7] Ji Junhong, Chang Runqi, Wen Tingxin. A Study on Traffic Accident Fatality Prediction Based on GSK-AdaBoost-LightGBM [J]. Safety and Environmental Engineering, 2021, 28(1): 24-28. DOI:10.13578/j.cnki.issn.1671-1556.2021.01.004.
[8] Yu Jiang. Research and Application of Data Mining Technologies in the Telecommunications Field [D]. Hunan: Xiangtan University, 2021.
[9] China Southern Power Grid Co., Ltd. and China Southern Power Grid Research Institute Co., Ltd. have developed a new renewable energy output prediction method based on the LGBM-Adaboost algorithm. The patent number for this prediction method is CN202410291938.7[P], with a publication date of May 7, 2024.
[10] Long Huifen. Analysis of User Churn Prediction in Music Streaming [D]. Guangxi: Guangxi Normal University, 2018. DOI:10.7666/d.D01508739.
[11] Lü Ning Luo Qian: A Telecommunications Customer Churn Prediction Model Integrating XGBoost and Logistic Regression Algorithms. School of Information and Communication Engineering, Beijing University of Information Science and Technology. 10.16652/j.issn.1004-373x.2025.11.021 2025-06-04
[12] Yan Chun, Zhang Xinyu: A Study on Life Insurance Customer Churn Prediction Algorithms Based on Improved K-means and BP-Adaboost School of Mathematics and Systems Science, Shandong University of Science and Technology 10.16452/j.cnki.sdkjzk.2022.01.006 2022-01-27
[13]Li Weikang, Yang Xiaobing: A Customer Churn Prediction Model Based on a Two-Layer Fusion Architecture, School of Information Engineering, China Jiliang University, September 9, 2020
[14] Zhou Jie, Yan Jianfeng, Yang Lu, Xia Peng, Wang Meng: Application of LSTM Model Ensemble Methods in Customer Churn Prediction
[15] School of Computer Science and Technology, Soochow University, November 29, 2019
Copyright © 2026 Ziyu Zeng, Maoxin Li, Yuyi Huang, Zhaohong Cao, Cunjue Li
Publishing time:2026-05-12
This work is licensed under a Creative Commons Attribution 4.0 International License