While machine learning (ML) techniques are widespread in pricing, they are not being adopted widely in reserving. In 2019, the IFoA formed the MLR WP, which includes more than 50 actuaries from around the world, to identify barriers, communicate benefits, and develop research techniques in pragmatic ways.
The scope includes:
- Find out where the market is on progressing ML
- Identify relevant papers
- Identify gaps
- Undertake research
- Look at specific questions around techniques, such as:
- What techniques should be reviewed
- Which techniques works best
- Incorporating actuarial judgement
- Measuring/spotting trends
- Quantifying uncertainty
- IBNER vs. IBNYR
- Using case estimates
- Using operational time/claims process
- Trust/ethics related to ML in reserving
- Identify data sources for testing
As the work is ongoing,the session will provide an update on the latest progress