I am a PhD candidate at the Department of Applied Physics, Stanford University. My research interest lies in the intersection between neuroscience, theoretical physics and computer science. Using machine learning methods and brain imaging data from mice, I aim to tap into the internal neural circuitry of the brain and understand the neural structure responsible for the memory formation and recollection in the short-term as a member of Schnitzerlab.

In my first year at Stanford, I joined an ongoing experiment to dissect the neural codes of short-term memory. I am analyzing data from a delayed cue discrimination task for mice, in which a well-trained mouse must form and use brief memories over the course of a few seconds — about the time that one might hold a phone number in mind. I train machine learning algorithms to reveal specific features of the neural data that subserve the memory in the cortex, which in turn leads to new hypotheses that my colleagues in the lab can test. My unique contribution to the project is searching for key patterns in the vast sea of big data using methods from graph theory, topology, optics and machine learning.

In my first few months in the lab, I have also realized that there is a widespread need to explain the types of computational tools that I use to biologist colleagues—both in my own and many other laboratories—in a pedagogical way. Thus, as an educational side project, I am creating a tutorial course entitled “Introductory Machine Learning Methods for Practitioners,” which aims to explain the most commonly used machine learning tools in research. By comparison, most prior resources of this kind are aimed at researchers seeking to advance AI research, whereas mine is aimed at practicing researchers. I plan to use my lecture notes to teach a class at Stanford aimed at practitioners, without making strong assumptions about their math background. These lecture notes will be available soon in this Github.