Description: Discussion of machine learning applications currently being used to augment traditional actuarial modeling in the life insurance industry. This includes data science techniques for: Quantifying and explaining differences between model versions or reporting periods Building proxy models for actuarial model output using machine learning One-class classification for data integrity checks Applications of unsupervised learning (e.g., clustering and principal components analysis) to reduce model run time

Description: We apply decision theory to develop a credibility formula that minimizes a loss function that is a linear combination of a squared-error term and a second-derivative term. The squared-error term measures the accuracy of the estimator, while the second-derivative term constrains the estimator to be close to linear.

Description: This presentation suggests a possible flexible solution to the time and resource problems of running a large number of stochastic interest rate scenarios, by selecting a representative subset. Each interest rate scenario consists of 30 future spot yield curves, where a reasonable number of points are specified on each curve such as 12. The distribution of the scenarios is approximated by the subset and each scenario in the subset has equal weight. The method is independent of the inter rate generator used.

Description: This is a presentation From 38th Annual Actuarial Research Conference 8/9/2003, University of Michigan, Ann Arbor, Michigan. Pricing health insurance presents numerous statistical problems. This paper develops a non-parametric method for handling these problems. This method involves kernel density estimation of the charge level distribution and also kernel smoothing between the reimbursement function derivatives. It derives some properties of the estimates so obtained. It illustrates the method with some dental data and compares the results to some other simpler estimation techniques.