Customer churn prediction based on the integration of meta-learning network of the forest
To address the challenge of capturing temporal features in customer churn prediction tasks by tree models, a churn prediction method based on ensemble forest meta-learning network (EFML) was proposed. Firstly, data quality was improved through grouping strategies and class imbalance issues were addr...
Saved in:
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | zho |
| Published: |
Beijing Xintong Media Co., Ltd
2024-10-01
|
| Series: | Dianxin kexue |
| Subjects: | |
| Online Access: | http://www.telecomsci.com/zh/article/doi/10.11959/j.issn.1000-0801.2024159/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1841528829251682304 |
|---|---|
| author | LI Longge ZHENG Kengcheng |
| author_facet | LI Longge ZHENG Kengcheng |
| author_sort | LI Longge |
| collection | DOAJ |
| description | To address the challenge of capturing temporal features in customer churn prediction tasks by tree models, a churn prediction method based on ensemble forest meta-learning network (EFML) was proposed. Firstly, data quality was improved through grouping strategies and class imbalance issues were addressed with undersampling techniques. Secondly, semantic vectors of user temporal features were constructed using EFML’s semantic graph constructor to depict fine-grained user behavior, forming a semantic graph and explicitly integrating it. Finally, multiple base tree models were trained as meta-learners, with the inputs being multilayer perceptron (MLP) to generate comprehensive churn prediction results. Experimental results demonstrate that EFML can effectively exploit differences in customer historical behaviors, capture and learn complementary relationships between base tree models. Compared to random forest (RF), EFML shows a 2.7% increase in AUC, a 3.7% increase in AP, and a significant improvement in prediction accuracy. This framework, combining tree models and micro-level features, possesses excellent interpretability, providing a new perspective for operators to achieve more refined user-centric management. |
| format | Article |
| id | doaj-art-1743799990904d82ae68439749a17b96 |
| institution | Kabale University |
| issn | 1000-0801 |
| language | zho |
| publishDate | 2024-10-01 |
| publisher | Beijing Xintong Media Co., Ltd |
| record_format | Article |
| series | Dianxin kexue |
| spelling | doaj-art-1743799990904d82ae68439749a17b962025-01-15T03:34:05ZzhoBeijing Xintong Media Co., LtdDianxin kexue1000-08012024-10-014016317276739101Customer churn prediction based on the integration of meta-learning network of the forestLI LonggeZHENG KengchengTo address the challenge of capturing temporal features in customer churn prediction tasks by tree models, a churn prediction method based on ensemble forest meta-learning network (EFML) was proposed. Firstly, data quality was improved through grouping strategies and class imbalance issues were addressed with undersampling techniques. Secondly, semantic vectors of user temporal features were constructed using EFML’s semantic graph constructor to depict fine-grained user behavior, forming a semantic graph and explicitly integrating it. Finally, multiple base tree models were trained as meta-learners, with the inputs being multilayer perceptron (MLP) to generate comprehensive churn prediction results. Experimental results demonstrate that EFML can effectively exploit differences in customer historical behaviors, capture and learn complementary relationships between base tree models. Compared to random forest (RF), EFML shows a 2.7% increase in AUC, a 3.7% increase in AP, and a significant improvement in prediction accuracy. This framework, combining tree models and micro-level features, possesses excellent interpretability, providing a new perspective for operators to achieve more refined user-centric management.http://www.telecomsci.com/zh/article/doi/10.11959/j.issn.1000-0801.2024159/semantic graphcustomer churnhistorical behavior differencetree model |
| spellingShingle | LI Longge ZHENG Kengcheng Customer churn prediction based on the integration of meta-learning network of the forest Dianxin kexue semantic graph customer churn historical behavior difference tree model |
| title | Customer churn prediction based on the integration of meta-learning network of the forest |
| title_full | Customer churn prediction based on the integration of meta-learning network of the forest |
| title_fullStr | Customer churn prediction based on the integration of meta-learning network of the forest |
| title_full_unstemmed | Customer churn prediction based on the integration of meta-learning network of the forest |
| title_short | Customer churn prediction based on the integration of meta-learning network of the forest |
| title_sort | customer churn prediction based on the integration of meta learning network of the forest |
| topic | semantic graph customer churn historical behavior difference tree model |
| url | http://www.telecomsci.com/zh/article/doi/10.11959/j.issn.1000-0801.2024159/ |
| work_keys_str_mv | AT lilongge customerchurnpredictionbasedontheintegrationofmetalearningnetworkoftheforest AT zhengkengcheng customerchurnpredictionbasedontheintegrationofmetalearningnetworkoftheforest |