Name: Anqi Wu
When: Thursday, December 3, 2020 at 11:00 am
Title: Understand The Brain Using Interpretable Machine Learning Models
Abstract: Computational neuroscience is a burgeoning field embracing exciting scientific questions, a deluge of data, an imperative demand for quantitative models, and a close affinity with artificial intelligence. These opportunities promote the advancement of data-driven machine learning methods to help neuroscientists deeply understand our brain. In particular, my work lies in such an interdisciplinary field and spans the development of both scientifically-motivated probabilistic modeling approaches for neural and behavior analyses and scalable and efficient inference algorithms to fit the models. In this talk, I will first present my work on a Bayesian latent manifold tuning model for neural recordings in multiple cortical areas and discovering intriguing scientific insights. Next, I will talk about a novel probabilistic graphical model for animal pose tracking and interpretable downstream behavioral analyses. In the third example, I will introduce a fast and robust inference method for Bayesian neural networks (BNNs), which are essential for studying higher-order sensory encoding. The proposed method eliminates gradient variance and automates prior selection, which are critical challenges in variational Bayes. Finally, I will envision my future plans on bridging these works and building integrated intelligent systems.
Bio: Anqi Wu is a Postdoctoral Research Fellow at the Center for Theoretical Neuroscience, the Zuckerman Mind Brain Behavior Institute, Columbia University, working with Prof. Liam Paninski and Prof. John Cunningham. She received her Ph.D. degree in Computational and Quantitative Neuroscience with Prof. Jonathan Pillow and a graduate certificate in Statistics and Machine Learning from Princeton University in 2019. She holds a Master's degree in Computer Science from The University of Southern California, and a Bachelor's degree in Electrical Engineering from Harbin Institute of Technology, China. She also worked as a summer research associate at The University of Texas at Austin and Microsoft Research, Cambridge, U.K.. Her research interest is to develop cutting edge machine learning models for in-depth scientific discovery in neuroscience and build up neuro-inspired artificial intelligence systems. Anqi received the USC Chevron Fellowship and was selected for the 2018 MIT Rising Star in EECS. She published a series of proceedings, and some of them were selected for oral presentations at top-tier machine learning and computational neuroscience conferences. She co-organized workshops at Cosyne and served as a Lecturer at Neuromatch Academy summer school.