Deep Learning for Cross-Selling Health Insurance Classification

Abstract

In the ever-evolving landscape of insurance services, the strategic cross-selling of health insurance to existing customers stands as a pivotal endeavor. This study delves into this domain, aiming to harness the potential of machine learning and deep learning methodologies to identify prospects for health insurance adoption within the existing customer base, irrespective of their current insurance holdings. To achieve this goal, this study conducts various experiments employing different machine learning and deep learning algorithms, such as TabPFN, Graph Convolutional Network (GCN), LightGBM, Hist Gradient Boosting and XGBoost. Data preprocessing involves filtering out empty data and normalizing the remaining data. The dataset is then subjected to various optimization methods, including Arithmetic Optimization Algorithm (AOA), Gradient-Based Optimization (GBO), and Sine Cosine Algorithm (SCA), to enhance the model's performance, ensuring the development of a reliable and trustworthy predictive model for identifying customers likely to purchase health insurance. Notably, AOA consistently demonstrates strong performance in LightGBM with an increment of 9.34%, XGBoost with an increment of 8.66%, and TabPFN with an increment of 10.48%, establishing it as the optimal method for enhancing model performance across different models. Moreover, an additional discovery in this study emphasizes TabPFN's notable superiority over other algorithms, showcasing a substantial improvement in accuracy from 52.62% to 63.10%. This substantial enhancement underscores TabPFN's exceptional efficacy specifically in managing tabular data.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 8 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.