Canadian Institute of Actuaries (CIA) has updated its 2017 educational notes. Using the modeladdressing the increasing use of machine learning (ML) models in property and victim (P&C) pricing.
“This educational note supplement is a review of methods and guidance regarding the use of the ML model in P&C pricing,” they write. Previous notes focused on traditional statistical models. The new update fills the gap in which new ML techniques emerged, they add.
“Many of the topics covered do not need to be extended in the context of ML models. For example, model risk, model selection, model limitations, documentation, and the use of models built by other models are relevant even if the model used is an ML model,” they stated in the note. Using machine learning models in P&C pricing. “However, the section on sensitivity testing, model implementation, data validation, and model validation require additional considerations not covered by the original educational notes. This supplement is intended to provide some insight into these additional considerations.”
Supplements contain discussions about definitions, use cases, and the general mechanisms of some common ML models.
“The actuarial profession has long used models for P&C insurance pricing and decision-making,” the paper states. “Updated guidance is needed for increased use of ML models in P&C pricing.”
