A few months ago, I sat on the first conference call with 43 other actuaries from around the U.S. to talk about how machine learning is changing the actuarial industry. The Casualty Actuarial Society (CAS) had formed a working party to research this area, and the turnout has been bigger than anyone has expected. We have since divided into two sub-groups to focus on 1) producing a worked-example of using ML, which is where I am, and 2) researching how to encorporate ML while following regulatory limitations.
The life-side of actuarial science, the SOA, also is highly active in this area, and they even have a monthly newsletter about predictive analytics which I would highly recommend to anyone interested See the SOA’s Section of Predictive Analytics and Futurism
This is a personal effort to gain familiarity with building models for pricing insurance products. Read the rough draft of my analysis here
Comments