News

Unpacking Data Insights: Lessons from the Scenario Response

Hello, my name is Omkar and I was a semifinalist in the 2022 Modeling the Future Challenge and a finalist in the 2023 Modeling the Future Challenge. 

The Scenario Response is a foray into the challenge that actuaries face, and the remainder of the MTFC. It consists of a set of prompts alongside a pertinent dataset. The prompts tackle risk and mitigation; analysis of risk both quantitatively and qualitatively is required. 

The first time I competed in the MTFC, the Scenario Response helped familiarize me with the idea of “cleaning” data. Cleaning refers to the act of removing incomplete rows, or more generally “bad” data. While it seems obvious, prior to MTF I had little experience with handling data, so the challenges involved in working with data were foreign to me. It has stuck with me, not just because it is a fundamental that is needed whenever one is working with actual data, but because it introduced me to interpolation. 

When data is removed, a “hole” forms in the data set. Among the many ways to address such holes is interpolation. Interpolation is a catch-all name for many such techniques. There are many enjoyable statistical challenges that are involved in interpolation and curve fitting. Good interpolation requires thorough knowledge of the “shape” of the data, and its context. Many interpolation techniques are accordingly related to extrapolation. 

Alongside “cleaning” data, the Scenario Response taught me to attach context to data. Prior to the MTFC, I only saw mathematics as connected to the real world in terms of physical laws. While I had a tangential awareness of math in relation to other concepts, such concepts were far from the forefront of my brain when thinking of math. However, the Scenario Response allowed me to view the context behind the data. It allowed me to see how math was a tool for analysis behind all sorts of data and phenomena. 

This has stuck with me because it is a fundamental idea behind any kind of risk analysis. For risk to be properly considered it needs to be quantified. We can always make general predictions about what will and won’t help, but quantification provides an idea of magnitude along with uncertainty. Uncertainty can be quantified by the means of probability and is a key component of risk. 

Overall, the Scenario Response is a small-scale guided version of the Project Phase that provides the much needed concepts and ideas behind the analysis of risk and actuarial tasks. It taught me the basics of working with data and opened the door to new exciting math. A thorough Scenario Response will teach those unfamiliar with MTF the basics they need to start with the project phase and provide those familiar with MTF practice. 

While I only spoke of cleaning data and mathematical contextualization, the project phase teaches a variety of other topics. Such topics include extrapolation, probability, expected value, and variance. All of these topics extend not only to the project phase, but also further into any other projects involving real data.