Description: Traditional risk scoring models include only fixed (non-random) demographic and condition-based variables. By including the provider effect as part of the risk scoring model as a random effect, we can get better estimates of provider performance (incorporating practice patterns conditional on patient morbidity profile). This approach can be generalized to study other factors that are now masked in error variance (i.e. area, Social Determinants) but if modeled could give us better estimates of risk

Description: This article provides a brief overview of the current state of machine learning (ML) and artificial intelligence (AI) as they apply to actuarial science, including discussion around the integration of actuarial judgment with ML and AI. It concludes with an example of explainable ML through a python tool known as SHAP (SHapley Additive exPlanations), which is able to explain the key drivers behind individual ML or AI predictions.

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: 1994 SOA Spring Meeting, Orlando. In this session the panel consider the question of uncertainty in loss reserves including discussion of sources and measurement. The panelists analyzed the same data set from different perspectives and discussed the strengths and weaknesses of the approaches.

Description: This chapter introduces a few probability distributions that play an important role in understanding and analyzing healthcare data. While theoretical in nature, these distributions model real life if appropriately chosen. Probability distributions are mathematical representations of how statistical data behave.

Description: While many know that machine learning is a big deal and even have a general idea of what it is, many do not know what machine learning can do for them. As an alternative to citing achievements or explaining algorithms, this article outlines the challenges that occur in statistics and the solutions that machine learning offers. Some of those challenges include correlated predictors, missing data, data interaction, and prediction intervals. Corresponding solutions include regularization technics, missing value imputation with KNN, tree based models, and cluster analysis for backtesting.

Description: Regression to the mean is a phenomenon that arises when comparing sequential data points. It states that when an initial observation is extreme or an outlier, future observations will be closer to the average. Health care actuaries need to understand and be able to mitigate the effects of regression to the mean to avoid misapplying predictive models or misstating measurements of historical effects.

Description: This chapter focuses on creating practical models that aid in the prediction of the risk of readmission to a specific hospital within 30 days of discharge for Medicare patients, to aid in the identification and stratification of at-risk patients for hospitals considering building an intervention program in a cost-effective manner. For readers who wish to develop their own predictive model, a dataset is provided on the Actexmadriver.com website, and R code is provided in the appendix to this chapter.

Description: Techniques that have gained some traction in general (property/casualty) insurance will be highlighted with a particular focus on the linear mixed model. This model adds random effects to the more commonly seen fixed effects (for example, as used in GLMs).

Description: This paper has two related purposes: first, to introduce a new class of Moving Weighted Average MWA graduation formulas where the coefficients are determined in accordance with the relative emphasis placed on smoothness and fit by the graduator, and second, to consider the question regarding what value of n to use when graduating by MWA formula. From the Actuarial Research Clearing House, 1978, Vol.1.