Download PDFOpen PDF in browser

Employing Machine Learning Techniques to Analyse Customer Records for Cross-Selling Probability

17 pagesPublished: June 16, 2024

Abstract

The study delved into health insurance cross-selling, where additional insurance products are promoted to existing policyholders, suggesting supplementary coverage such as dental or life insurance to those with basic health insurance. The study focused on applying machine learning to predict cross-selling opportunities among South African customers. The aim was to develop a predictive model to aid health insurers in identifying potential cross-selling customers. Utilizing a quantitative research methodology, a comprehensive dataset of health insurance consumer information was analyzed using various machine learning algorithms, including Random Forest, K-Nearest Neighbors, XGBoost classifier, and Logistic Regression in Python. Results revealed Logistic Regression as the top-performing model, achieving an accuracy score of 0.83 and an F1 score of 0.91 when trained on a dataset of 1,000,000 health insurance customers with 17 features comprising health insurance customer information. The analysis uncovered that customers aged 25-70 with prior insurance and longer service history are more likely to purchase additional health insurance products. These insights empower health insurers to enhance revenue through improved customer targeting and retention strategies, thereby providing valuable information for the industry’s understanding of effective cross-selling approaches. The methodology comprised quantitative data extraction and machine learning application, thus contributing to advancements in cross-selling strategy comprehension.

Keyphrases: cross selling, health insurance customer records, machine learning, model evaluation matrix, prediction

In: Hossana Twinomurinzi, Nkosikhona Theoren Msweli, Sibukele Gumbo, Tendani Mawela, Emmanuel Mtsweni, Peter Mkhize and Ernest Mnkandla (editors). Proceedings of the NEMISA Digital Skills Summit and Colloquium 2024, vol 6, pages 78-94.

BibTeX entry
@inproceedings{NEMISADigitalSkills2024:Employing_Machine_Learning_Techniques,
  author    = {Khulekani Mavundla and Dr Surendra Thakur},
  title     = {Employing Machine Learning Techniques to Analyse Customer Records for Cross-Selling Probability},
  booktitle = {Proceedings of the NEMISA Digital Skills Summit and Colloquium 2024},
  editor    = {Hossana Twinomurinzi and Nkosikhona Theoren Msweli and Sibukele Gumbo and Tendani Mawela and Emmanuel Mtsweni and Peter Mkhize and Ernest Mnkandla},
  series    = {EPiC Series in Education Science},
  volume    = {6},
  publisher = {EasyChair},
  bibsource = {EasyChair, https://easychair.org},
  issn      = {2516-2306},
  url       = {/publications/paper/dm86},
  doi       = {10.29007/rs1z},
  pages     = {78-94},
  year      = {2024}}
Download PDFOpen PDF in browser