Integrating Peak Ground Acceleration as a Damage Factor in Risk-Based Premium Rate Assessment using K-medoids Bayesian networks

Abstract

In the insurance market, determining fair and acceptable premium rates requires an accurate evaluation of risk. In the context of earthquake damage, the Peak Ground Acceleration (PGA) level is essential for assessing the intensity of ground shaking and its effect on structures. However, the present approaches for adding the PGA level as a damage factor in risk-based premium rate calculation are inaccurate and inefficient. This study proposes integrating the PGA level as a damage factor using Bayesian networks to overcome this issue. Using the probabilistic nature of Bayesian networks, the suggested solution provides a more complete and accurate method for determining premium rates. The premise is that the integrated Bayesian network model will produce more accurate calculations of premium rates than previous techniques. This work is significant because it has the potential to improve the fairness and openness of premium rate determination, resulting in enhanced risk assessment methods in the insurance business. By taking into account the unique impact of the PGA level on building damage, insurers can better align premium rates with the real risk profile of insured items, which is advantageous for both insurers and policyholders. According to the research findings, the premium rate increases as the level of risk in a location rises. Incorporating PGA and the extent of damage, the output of the BN model can also be used to estimate the premium rate per subdistrict. This analysis clearly demonstrates that the premium rates varied by subdistrict.

Keywords

• earthquake
• building damage
• bayesian network
• k-medoids