Kevin F
- Research Program Mentor
PhD candidate at Stanford University
Expertise
Machine Learning, Data Science, Statistical inference.
Bio
Hi! I am a second-year PhD student in Statistics at Stanford University. In between my MS and starting my PhD, I worked at BlackRock in their AI Labs group working to apply statistics, optimization, and machine learning to problems throughout the firm. My main interests are in modern statistical inference methods such as selective inference and conformal inference, with an eye towards developing and maintaining software for such methods. When I'm not doing those things, you'll often find me bouldering, hiking, or learning a new recipe.Project ideas
Machine Learning Application
This end-goal of this project would be to build a predictive machine learning model on a dataset of your choice (or one we can find together). If you have no prior experience with machine learning, we would begin by going over the fundamental models and concepts of the field before moving on to the application. If you already have a sufficient background, we can dive straight into the data analysis.
Understanding the "Reproducibility Crisis"
In this project we would go over several important readings and papers to understand the so-called "reproducibility crisis" in science. "Reproducibility crisis" refers to there being a large number of studies whose findings cannot be reproduced by other researchers. For instance, one psychology report was only able to reproduce 36 out of 100 findings from top journals. We will begin by discussing the statistical, game-theoretic, and structural reasons causing the crisis, and then discuss several new tools scientists have developed to overcome it. The culmination of this project could be a blog post summarizing what we have learned and suggestions for further improvement. Alternatively, a student could also conduct an analysis of these tools to compare and contrast their efficacy.