BYU

Abstract by Jared Cummings

Personal Infomation


Presenter's Name

Jared Cummings

Degree Level

Undergraduate

Abstract Infomation


Department

Statistics

Faculty Advisor

Brian Hartman
Shannon Tass

Title

Using Deep Learning to Forecast Long-term Care Insurance Claims

Abstract

One of the most important things that actuaries do is predict the incidence of future claims. This is especially important in long-tailed liabilities like long-term care insurance (LTCI). We will develop new deep learning methods to predict LTCI claim incidence and compare the results to the current state-of-the-art in LTCI claim forecasting.  In addition to the new deep learning methodology, we will compare and contrast several different model selection methods to help industry actuaries apply these models to their work.