Meet a Data Scientist: Alex Franks, PhD
17 January 2023
BY STELLA MIN
Thanks to the fantastic work of Tiana Curry, our former Community and Communications Assistant and current grad student at the USC Viterbi School of Engineering, we’re featuring an interview with CDN member Dr. Alex Franks this month. During the interview, Dr. Franks shared what drew him to statistics and data science and tips for collaborating with subject matter experts.
Dr. Franks earned his Bachelor's in Computer Science and Applied Math from Brown University. He earned his PhD in Statistics from Harvard University, where he focused on biological applications under the supervision of Dr. Edoardo Airoldi. He then completed a postdoc at the University of Washington with Dr. Peter Hoff.
Initially, Dr. Franks was drawn to the field of mathematics because of his interest in robotics. However, he chose to study statistics because of its applicability across many different fields and industries. Needless to say, he does not regret his decision.
Dr. Franks is an Assistant Professor in the Department of Statistics and Applied Probability at the University of California, Santa Barbara. His research interests include covariance estimation, multivariate analysis, high dimensional data, errors-in-variables models, causal inference, missing data, and spatiotemporal methods. He works on applications in computational biology and sports analytics.
Because Dr. Franks specializes in statistics, he often collaborates with subject matter experts in biology. His approach to navigating these research collaborations is to ask a lot of questions to understand a particular measure, why there might be missing data, and why processes or variables may be correlated. He then abstracts away from the biological processes to form a statistical model, after conducting a thorough exploratory analysis.
Dr. Franks encourages his students to adopt a similar approach when learning statistics. That’s because, unlike mathematics, there’s no right answer in statistics. Statisticians can choose to model an underlying process in different ways, depending on hypotheses and assumptions about the data and underlying processes. Dr. Franks acknowledged that this is often confusing for students since statistics classes are often taught like ones in math.
Outside of Work
How to Connect
Check out Dr. Franks’ profile in the CDN Member Directory which shares a link to his social accounts, GitHub, and website.