Life insurance model with regression cox proportional hazard affected by areal spatial factor

Open

Lita Wulandari Aeli, Andi Daniah Pahrany, Sapto Wahyu Indratno

2024 AIP Conference Proceedings Vol. 3049 Issue 1 Conference paper Cited by 0 Quartile

Abstract

Life insurance discusses about financial self-protection against uncertain events by focusses the state of the area of residence in addition to commonly used demographic information, such as age, gender, occupation, knowing certain diseases and others. Areal data contains postal codes, district/cities and regions. The application of the Weibull distribution to Cox Proportional Hazard regression can be used to describe the death and survival function for individuals living in a certain area by inputting spatial frailty effect. Spatial random effects on areal data are represented by a conditional autoregressive distribution that pay attention to the neighboring relationship of an area. The estimation of parameter values from the initial model can be obtained by using Maximum Likelihood Estimation (MLE) and continuing with the Newton-Raphson optimization steps. The results of life insurance premiums without regard to random effects and with regard to random effects were simulated and applied to case data of acute myeloid leukemia in adults in North West England between 1982 and 1998 by the North West Leukemia Register in UK. © 2024 Author(s).

Affiliations

Universitas Negeri Malang, Malang, Indonesia; Institut Teknologi Bandung, Bandung, Indonesia