Exclusion Vis: A Visual Interactive System for Exclusion Analysis in Life Insurance Using Machine Learning Techniques

Exclusion analysis in life insurance stands out as a crucial task in minimizing customer risks. Numerous insur-ance companies have implemented artificial intelligence (AI) solutions to streamline processes and mitigate customer-related risks, specifically leveraging machine learning (ML) technolo-gies. While some studies have acknowledged its significance, there needs to be more to deliver comprehensive research on exclusion analysis utilizing advanced techniques. Thus, this paper focuses on exclusion classification blending with visual interactive systems (Vis), aiming to comprehend data obtained from customer disclosure information. To tackle this challenge, 1) we initiate the investigation by employing various multi-label classifiers, 2) apply several ML techniques, utilizing data from a prominent Australian insurance company, where support vector classification (SVC) and random forest (RF) emerges as the superior performer, and 3) we introduce a visual interactive system called Exclusion Vis, meticulously crafted to facilitate in-depth exploration of the obtained results. The evaluation of Exclusion Vis further underscores its superiority compared to existing systems.