Description: This article discusses how to handle semisupervised data. This type of data might have a large number of observations, yet a small number of labels.

Description: Data needed to run the Hack-a-Thon that was held at the 2019 PAS.

Description: Session 2F: Artificial Intelligence and Actuarial Modeling - Ask an Expert!

Description: Modern Actuarial work, being quite data intensive, provides many opportunities for actuaries to use the statistical skills that are part of their basic training. Furthermore, modern computing allows us to analyze vast quantities of data intensively, and each package will provide statistical 'answers' to any questions the actuary might ask of it. However, in this particular field, it is often worth taking a step back, and ensuring that we recall and consider the warnings that come with each of the statistical procedures we might attempt to use. This will allow us to ensure that we guard ourselves against drawing inappropriate conclusions, which could come back to haunt us later.At the conclusion of the session, attendees will be able to avoid drawing inappropriate statistical conclusions, while doing their work. They will leave informed about the types of pitfalls that exist, and where those might come into play in actuarial work.

Description: En el seminario hablaremos sobre los nuevos flujos de tarificación, y nos enfocaremos en la inclusión del análisis geográfico en el proceso de modelado para optimizar la extracción del riesgo en modelos de Machine Learning.

Description: This session will cover everything you've wanted to know about GLMs but were too afraid to ask. How do these generalized models relate to the plain linear regression, and how can we be sure that the models we implement are appropriate for the task at hand?

Description: Health claims are not normally distributed. Claims Forecasting can be improved by understanding the entire distribution of plausible values of future claims, for example knowing there is an 80% chance future claims are between $340 and $375 PMPM. Confidence intervals around our estimates allow us to understand for each range around the best estimate (for example +/- 1%) what is the probability that future claims will fall in that range. Quantile Regression is a mathematical technique where you can estimate any percentile of the claims distribution you are looking for as a function of demographics, diagnosis indicators, prior cost variables, or other significant predictors. You can see which coefficients are important at different parts of the distribution. Individual and group distributions can be constructed. There are numerous applications in pricing and underwriting, stop loss, provider profiling, ACO target setting, allocation of group risk charges based on volatility, discounting strategy, and experience analysis.

Description: Python is the leading programming language for data science. Python programming is one skill that actuaries should acquire. It facilitates analytic work, and the available visualization tools help to communicate ideas. In this article, I use a traditional actuarial problem, mortality modeling, as an example.

Description: In How Significant Is Statistical Significance? The author discusses the p-value that is often used to determine significance. Sometimes we place too much reliance upon canned metrics and lose the ability to notice when something is mathematically sound but not sensible. The article provides three simple examples that require no complicated mathematics to follow yet provide a basis for some deep thought about statistical significance versus substantive significance.

Description: The presenters will explore policyholder experience data for long-term products, risks of over-fitting data in developing behavioral models, and ways to mitigate these risks and develop a sustainable and coherent framework to translate complex policyholder behavior experience data to behavioral models using company- and industry-level data. At the conclusion of the session, attendees will be able to: - Identify key drivers of policyholder behavior for long-term product - Understand the risks of over-fitting data and a false sense of precision in developing behavioral mode - Understand techniques that can be used to mitigate these risk