Description: Actuarial concepts of credibility can be extended to encompass additional information provided by new machine learning estimates. New statistical ideas also indicate how to generalize to blend arbitrarily many estimators, while not double-counting common information. For example, historical claim costs and risk scores both leverage diagnosis code information. We introduce two new estimators, one based on partial correlation and the other based on the machine learning-derived Shapley statistic, to address these limitations.

Description: The July 2013 issue of the SOA’s Forecasting & Futurism section’s newsletter, Forecasting & Futurism.

Description: Feature article discussing model monitoring

Description: An exploration of moving predictive modeling capabilities in R to real time operations.

Description: Reviewing a text covering multiple statistical modeling topics: e.g., generalized linear models, decision trees, clustering, using the statistical language R.

Description: This story won the Predictive Analytics and Futurism Section prize for most novel prediction forming the basis for the narrative in the 2019 Actuarial Speculative Fiction contest.

Description: December 2016 edition of Predictive Analytics & Futurism

Description: Announces author's departure as Section Chair and incoming Council members.

Description: Uncertainty of a predictive model is a fact of life that many insurers could be overlooking at their peril without a framework for assessing it. The article provides a framework to understand the prediction uncertainty, which can be critical for the insurance operations using predictive models. Coin toss example explains model uncertainty and randomness inherent in making predictions to the number of heads expected. A mortality analysis case study illustrates how the Bayesian methodology can be used to develop the prediction uncertainty. In particular, the article shows that the Bayesian framework is a way to combine historical data and actuarial judgment, and directly address the model uncertainty through its posterior distribution.

Description: Machine learning models are not black-boxes; these models can easily be interpreted using tools like SHAP, LIME and Anchor. These tools can facilitate the understanding of the model and the business problem. We can used these tools to build better models and communicate how predictions are generated. This article demonstrates their use in a mortality model.