MStat Seminar

We'll explain what students are expected to do in the project and how the process usually unfolds. We plan to have a mock defense from a previous student so you can see how the defense typically works, and of course there will be time to answer your questions.

Robert Cator is an actuary based at Aegon Insurance in the Netherlands. Aegon is one of the largest issuers of life insurance in Holland, and in America through its subsidiary Transamerica, Robert will tell us about the work he does as an actuary, the types of challenges that they face in setting prices for insurance contracts, and the data they use to make their analysis.

As the rate of data generation continues to grow, the ability to utilize quantitative methods to inform business decisions is increasingly valuable. But how does statistics fit into this new world where sample sizes are large, optimization problems are not convex, and high dimensional data take the form of images or text? In this talk, I argue that statistics is the foundation of many applications seen at the forefront of today's data renaissance. At the same time, there are many tools outside of statistics that are needed to take full advantage of modern data. Using examples from my work as a data scientist at Recursion Pharmaceuticals, I discuss the importance of statistical concepts such as maximum likelihood estimation, experimental design, and hypothesis testing. I then show how these concepts inform and relate to machine learning methods such as semi-supervised convolutional neural networks, stochastic gradient descent, and prediction in the context of biological images.