Description: A description of Data Science best practice tools and how they can apply to general actuarial work.

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: At the 2009 SOA Annual Meeting in Boston, Nassim Nicholas Taleb shared some thoughts on Black Swans, operating in the fourth quadrant, and living in the world of Extremistan versus Mediocristan during his keynote speech, all of which can be found in his published work. The follow-up session, summarized in this article, addresses the real question: What do we actuaries do about these phenomena?

Description: In this article we examine the benefits and drawbacks of using a global surrogate of a black-box model. Are surrogates a comfortable middle ground that provide insurers high predictive power as well as model explainability?

Description: This article provides a brief introduction to forecasting methods that are potentially relevant to actuarial work.

Description: Read the December 2018 issue of Predictive Analytics and Futurism published by the Predictive Analytics and Futurism Section.

Description: A significant fraction of insurance risk is characterized by the tail of the claims distribution, where we have much less data to model probabilities accurately. Ignoring the tails can lead to underpricing of tail risk and mispricing. Estimation of the tails can be complex due to a bias-variance tradeoff--the further out we go into the tail the more accurately we can model the extreme tail (lower bias) but the less data we have (increased variance). The statistical tools of Extreme Value Theory can be leveraged to more accurately estimate both the incidence and the severity of extreme events given that they occur. using the Generalized Pareto Distribution.

Description: Read the April 2019 issue of Predictive Analytics & Futurism published by the Predictive Analytics & Futurism Section of the Society of Actuaries.

Description: Given the ubiquity of judgmental forecasting, it is important to understand the potential biases inherent in human judgment—the traps that can snare us—and to learn methods and best practices to avoid them. The aim of this article is to help readers with these goals.

Description: The author explains how he used judgmental forecasting techniques to develop a dynamic lapse assumption. He defines the techniques and provides readers with a step-by-step process.