What is your specialty?
I am a computational cognitive neuroscientist. I’m really interested in brain network dynamics and larger systems that relate to patients where you can pick out biomarkers that are similar across patients to help diagnoses.
How did you end up researching the brain?
I started as a studio art major and a chemistry minor. I thought I would go into medicine, but I took a gap year and worked on a dude ranch, and I tried my hand in New York City as an artist for a while, all while thinking that I would go back to med school. But then I started tutoring in math and science, and I really fell in love with teaching. I got really interested in how people learned and remember and what made for efficient and effective teaching. I enrolled in a master’s of neuroscience and education at Columbia University and that was my avenue into research. It was there that I got more and more interested in the brain specifically, which led me to pursue my PhD at Dartmouth College.
How did you get into computer science?
I didn’t start coding until I was past my master’s. I took to coding really quickly. It just made sense to me. [Coding] is a very interesting optimization problem. You've got something that you need to get done on the computer, and there are a million different ways to do it, how do you do it elegantly and efficiently?
My PhD research focused on modeling brain data at high temporal and high spatial resolutions, and I got more and more enmeshed in the computer science aspect of it. It also honed my interest in machine learning and software design.
Did your post doc focus on the same subject?
No, it was different. I have always had a strong interest in patient outcomes and translational research that applies to clinical applications. For my postdoc at Brown we were working with chronic pain, which is a weird disease in that there can be many different causes. There are so many different acute pain states that lead to chronic pain. And there are so many different manifestations of chronic pain.
We were trying to figure out if there were biomarkers could predict whether a person was going to develop chronic pain using reinforcement learning, specifically avoidance learning.
We were working on a paradigm for people that they could interface with on their
phone or computer, and then the differences in the in the way that they participated in that reinforcement learning game could help predict whether or not they were likely to develop chronic pain.
How will your expertise apply to SMART FIRES?
As a cognitive neuroscientist, I am familiar with ways in which we can use sparse and missing data to make predictions. I have experience with several machine learning techniques that I think could be useful in fire mapping.
Specifically, I have used Gaussian processes in several of my projects. For my graduate work I used Gaussian process regression to predict whole brain activity from a subset of electrode locations by leveraging neural similarities across patients.
My work has shown how you can use assumptions like spatial smoothness and covariance across people to make accurate predictions about missing data. Additionally, for my internship at Facebook Reality Labs we used Gaussian processes for adaptive psychophysics experimentation. Although the goal for the project was to develop faster and more efficient ways of performing human psychophysics experiments, I think these tools and ML technique like them might be useful in prediction and prevention of wildfires in several ways.
I think these tools could help predict the spread of wildfires by developing a model of factors that contribute to wildfires. Another way would be to leverage the uncertainty in the data to best inform sensor placement to help in the prevention of wildfires. Additionally, generational changes to forests due to fires and other long- term changes would be interesting to explore. Understanding the relationship between these variables and modeling them across time is something well suited to my previous work in timeseries modeling.
What are you most excited about with your new position?
I really like to teach. Specifically, I find adapting to so many different students and types of learning to be a rewarding challenge. The neuroscience of learning is how I got into my field in the first place.
What do you do outside of work?
Hiking, reading, gardening, and painting.