Description: ALM and hedging can reap significant benefits from innovation in computation and operations research. Explore applications of static and dynamic optimization in two different ALM and hedging use cases. Insurers are increasingly moving offshore by establishing Bermuda captive reinsurers. In calculating the Bermuda Best Estimate Liability ('BEL'), reinsurers will need to segment assets to support liabilities in each treaty. A Linear Programming ('LP') problem can be formulated to tactically select assets in a way that minimizes the BEL and consequently the total asset requirement. Learn more about what US Variable Annuity Insurers face with the regulatory requirement to reflect 'future hedging strategy' in their statutory funding calculation. The traditional numerical method of finding the optimal hedging target and then allocating assets may encounter the curse of dimensionality with expanding product and economic environment complexity. A Reinforcement Learning algorithm may well fit this problem and provide insurers a solution for dynamic hedging.

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: This article gives a brief overview of ways to extend clustering analysis beyond basic methods such a k-means

Description: CMS has required hospitals to publish their chargemaster rates, cash discounts and negotiated rates as of January 1, 2021. Our panel is comprised of a data strategist, an executive from a large hospital system and an actuary.

Description: This session is a primer on how to apply Gestalt psychology to display quantitative information to improve neural receptiveness. While much of brain functioning is still a mystery to psychologists and neuroscientists, we do understand a great deal about which parts of the brain engage or disengage given visual cues. Actuaries are often tasked with presenting very technical information and are prone to focus more on the substance of numerical data rather than a pictorial presentation based on neural preferences. This presentation will discuss how to display quantitative information more effectively tailored to these preferences.

Description: "Explore a general introduction to the opportunity of applying a relatively newer area in machine learning called uplift modeling, which has been successfully used in retail (Wayfair etc). The concepts are parallel to the intervention programs in health care, where we are interested in measuring the treatment effects and ultimately directing the right members to the most effective programs, driving the highest value for the programs both from a financial and health outcomes perspective. Uplift modeling can also be viewed in context of causal modeling where we are interested in assigning cause to a treatment relative to a member not receiving the treatment. Check out an introduction on the concept, seehow it relates and builds on existing work, and use a demo to showcase the potential of integrating this ML approach to current work."

Description: This session will examine bias from a social science perspective. It will discuss the three types of systemic biases using the National Institute of Standards and Technology (NIST) model. The NIST model details the three types as historical bias, societal bias, and institutionalbias. The session will also provide several examples of each type to provide a framework for identifying the social structure of bias.

Description: "Discuss research conducted at the Cleveland Clinic to understand levels of anxiety and depression across Northeast Ohio. Specifically, learn how advanced analytical techniques were utilized with social media data from Twitter and Google to estimate the need for psychological support by geography and overtime. The panel will go further to demonstrate how natural language processing (NLP) and machine learning could be used to make operational decisions at a leading health care provider."

Description: Model design is an important part of being an actuary, and the SOA highlights the importance of the actuarial control cycle of model design, evaluation and monitoring. What tools are at our disposal for these steps? When should we be concerned or excited about model performance? Can we, and should we, use machine learning tools to assist with model improvement or development?

Description: Analyze the market volatility behavior in context with historical extreme events during this webcast. There will be a focus on understanding volatility clustering and relationships, both contemporary and temporal, in addition to key economic and capital market variables. You'll also learn from a comprehensive attribution analysis, which uses economic data, asset data, retail investor data and event data as explanatory variables. Dive into investment-related Reddit comments used to capture features of retail investors and tree-based models, including Random Forests and Gradient Boosting Machines to justify changes in volatility modeling. Expect this webcast to end with an example of liability-driven investment optimization that illustrates the potential impact of increased volatility assumption, how to adopt stochastic volatility models, how to reflect volatility clustering and return jumps.