Life After CDS: A Conversation with CDS Faculty Fellow Angela Radulescu
As CDS continues to grow, so do our people. You may remember CDS Faculty Fellow Angela Radulescu from her guest editorial piece, can data include personal narrative? featured on the blog last spring. We’re thrilled to announce that she will start a position as an assistant professor of psychiatry and principal investigator at the Center of Computational Psychiatry at Mount Sinai’s Icahn School of Medicine. We caught up with Angela to discuss her experience at CDS, her new role, and how she hopes her research will impact the future of data science.
This interview was lightly edited for clarity.
What initially brought you to CDS?
I received my PhD in cognitive psychology from Princeton University, where my dissertation was focused on human reinforcement learning. I joined CDS in order to gain more exposure to modern methods in deep learning and NLP that, when combined with reinforcement learning, can provide theoretical accounts for how humans are able to learn efficiently in multidimensional environments.
Can you tell us a bit about what you’ve been working on at CDS and describe your overall experience here so far?
I’ve really enjoyed the intellectual independence that comes with being a CDS Faculty Fellow. I spent my time here developing a few new lines of work studying reinforcement learning and decision-making in naturalistic environments. In one project, my collaborators and I are asking how humans represent structure in virtual reality scenes — what are the types of priors that drive different decisions (e.g. where to look for objects), and how can we formalize them in a computational model of gaze behavior? In another project, I am looking at whether language priors influence how people learn in multidimensional environments. Being exposed to the diversity of ideas and approaches at CDS was exciting and helped me learn a lot about different areas of ML in a short period of time. I also felt very supported in endeavors that prepared me well for my next position such as grant writing and management.
You’ll be starting a new role. Can you tell us a bit about your new position and what you are most looking forward to working on there (as it relates to data science)?
I’m most excited about taking the work I’ve been doing in computational modeling and naturalistic reinforcement learning and applying it to understanding how learning can lead to changes in mental health. Computational psychiatry is an interdisciplinary field that blends a lot of core data science methods (e.g. reinforcement learning, inference, supervised learning, NLP, and more) to answer questions of relevance to understanding how mental health changes over time. The long-term hope is that we can take some of that understanding and apply it to improving treatment.
What role do you see your work/research potentially playing in the future of data science?
I think human cognition is a really important source of inspiration for some of the most exciting work going on in data science. In recent years, there have been a lot of great examples of the virtuous cycle of research between the study of the human mind and artificial intelligence. The new Minds, Brains and Machines initiative at NYU really embodies that ethos. I see my future research as a bridge between these evolving ways to think about the human mind and using that knowledge in clinically relevant contexts.
Do you have any final thoughts or comments you’d like to share about yourself, your work, CDS, or data science in general?
I’m very grateful for both the professional and personal connections I’ve made at CDS and at NYU more generally. In the broader context of the last couple of years, it’s been really important to be part of a department where I felt supported and surrounded by colleagues who were both genuinely invested in my success, and provided great examples of kindness and resilience during a difficult time for all.
By Ashley C. McDonald
