This analysis proposes a novel Gaussian process regression approach to enhance claims reserving in actuarial sciences, indicating potential improvements over traditional methods.
Key Points
The goal is to explore Gaussian process regression models for claims reserving and their effectiveness in actuarial science.
Utilized Gaussian process regression for modeling claim amounts.
Introduced an inverse Wishart distribution approach in actuarial modeling.
Employed open-source software and NAIC Schedule P dataset for modeling and data collection.
Provided comparisons with traditional reserving methodologies.
Demonstrated the ability of GPR to provide predictions with uncertainty intervals.
Highlighted improvements in capturing correlations between observed and expected claims.
Showed the potential of GPR techniques to extend traditional stochastic reserving methods.