September 9, 2019
Dr. Stephanie FitchettData Scientist, Transamerica
Come and meet with Dr. Fitchett
Machine learning models are becoming ubiquitous, as are data science positions for those who build them. This talk will start with a general discussion of machine learning, then focus on some specific machine learning models under development for life insurance underwriting. We will discuss how one or two models work at a high level, some of the challenges in building useful models, and the safeguards that are in place to prevent unintended negative consequences. If there is interest, we can also talk about data science as a career path.
There are many opportunities in industry for students with strong statistical and mathematical backgrounds—not necessarily majors. This talk will share a handful of projects that have a statistical or mathematical component and are driven by science or business needs. The projects include tracking the effects of landfill contaminants on fish over time; detecting anomalies in “pattern or life” transportation routes using imagery from synthetic aperture radar; estimating the extent of music piracy among an internet provider’s customers; and modeling the spread of avian influenza in Southeast Asia.
Stephanie Fitchett holds a PhD in mathematics from the University of Nebraska and an MS in statistics from Colorado State University. Following a postdoctoral position, she spent 12 years as a faculty member at Florida Atlantic University’s Honors College and at the University of Northern Colorado. She has worked as a statistician in industry for the past seven years, first at Sandia National Laboratories, and later at Neptune, a small environmental consulting firm. In her current position as a data scientist at Transamerica, a life insurance company, she builds statistical models that leverage corporate and external data to facilitate underwriting of life insurance policies. Steph lives in Boulder, Colorado, and is an avid whitewater kayaker.