Current Position
I am a Postdoctoral Fellow in Mathematics at the University of Victoria, BC, Canada (UTC-8).
Current Position
I am a Postdoctoral Fellow in Mathematics at the University of Victoria, BC, Canada (UTC-8).
Research Overview
My research lies at the intersection of Applied Topology and Neuroscience, focusing on the geometry underlying neural activity. My goal is to develop predictive biomarkers that enhance diagnosis and treatment selection for neurological and psychiatric disorders, with a particular emphasis on migraine patients.
I analyze dynamical brain scans (fMRI) using methods inspired by Persistent Homology, Network Analysis, and Knot Theory. The resulting invariants are then used to train Machine Learning models for predictive analysis.
Currently, I collaborate with Stanford's Brain Dynamics Lab (Prof. Saggar & Dr. Siu) and the Italian National Institute of Health (Prof. Branchi & Dr. Viglione) to apply these methods in biomedical research.
Previous Work in Algebraic Topology
Before transitioning to Topological Data Analysis (TDA), my Ph.D. research focused on knot theory through the lens of Goodwillie-Weiss Calculus. By sampling a knot with an increasing number of points, one obtains an approximation of the space of all knots through a discrete sequence of finite configurations. This process gives rise to the Goodwillie-Sinha Spectral Sequence, that provides precious information on the cohomology of the space of all knots.
My contribution, in collaboration with Prof. Salvatore, involved studying this spectral sequence modulo 2, resolving a conjecture by Vassiliev. As zero-dimensional cohomology of the space of all knots describes knot invariants, these calculations have the potential to shed light on classical problems from knot theory.
In parallel, I am working with Dr. Casarin on infinity-categorical approaches to multicomplexes—a homotopic deformation of bicomplexes that enables significant simplifications in spectral calculations. We plan to extend the multicomplex-based methods from my Ph.D. thesis, developed for the Goodwillie-Sinha Spectral Sequence, to a broader class of Bousfeld-Kan spectral sequences.
Collaboration & Contact
I welcome collaborations, particularly in applying TDA and ML to biomedical data science. If you’re interested in working together, feel free to reach out at:
andreamarino [at] uvic [dot] ca