Topology-Driven Insights into Naturalistic Behavior from Neuromechanical Agent Modeling

Cover page video in courtesy of Donald R. Sheehy's group.

We like to talk about and quantify things in Euclidean space where our intuition works the best. However, most of the times it might be complicated trying to find the correct "metric" of understanding things here. Under such a structured space, we sometimes don't know how to talk about complex, yet interesting, data. Computational Topological Data Analysis (TDA) operates on a combinatorial space, which may provide us understandings of the shape of data in a more abstract but comprehensive way. I am currently utilizing TDA to understand the representations of naturalistic behaviors learned by neuromechanical agent in Mimic-MJX with advising from professor Yusu Wang and in collaboration with the VNL team in Talmo's Lab. Latest rport on the progress can be found in this end of quarter presentation.

TDA
Our TDA pipeline for behavior segmentation.

Codebase (closed repo now but will be open later)