About

I am a post-doctoral researcher at UPenn working with Rob Ghrist. I received my PhD in Mathematics from Rutgers in October 2012 under the direction of Konstantin Mischaikow. Before coming to Rutgers, I earned a Bachelor's degree in Computer Engineering as well as a Master's degree in Mathematics at Georgia Tech. Here is a brief CV. You can also find me on MathOverflow, on Google scholar and on LinkedIn.

me

Contact

Mon, Wed, Fri @ Penn
3C5 David Rittenhouse Lab
209 South 33rd Street,
Philadelphia, PA 19104

Tue, Thu @ Rutgers
269 Hill Center
110 Frelinghuysen Rd,
Piscataway, NJ 08854

Email: penn or rutgers

Research Focus

I develop theories, algorithms and software for algebraic-topological analysis of large datasets which arise from simulations or measurements made in a variety of contexts.

Recent Projects

Discrete Morse Theory for Persistence

Discrete Morse theory is a wonderful tool from combinatorial algebraic topology which transports Morse theory from smooth manifolds to the realm of cell complexes. Critical cells of discrete Morse functions on a given complex generate a new (and smaller) complex possessing the same homology groups. We extend discrete Morse theory to filtrations of complexes and provide a novel approach to efficient computation of persistent homology. Here is an online version of the article.

The Perseus software project implements these ideas to compute persistent homology very quickly.

This is joint work with Konstantin Mischaikow.

Reconstructing Functions from Dense Samples

Given a compact Riemannian submanifold of Euclidean space, it is possible to recover (with high confidence) the homotopy type of this manifold from a sufficiently dense uniform point sample. In our work, we demonstrate a similar probabilistic result for functions. We show that under mild assumptions it is possible to reconstruct a continuous function between two such manifolds up to homotopy if we are given: a) dense point samples taken from the domain and the co-domain, and b) the images of the domain samples under the action of the function. The result is robust under perturbations arising from bounded sampling noise. Here is a preprint of the result.

Joint work with Konstantin Mischaikow and Steve Ferry.

Topologically Measuring Protein Compressibility

Joint work with Marcio Gameiro, Yasuaki Hiraoka, Shunsuke Izumi, Miroslav Kramar and Konstantin Mischaikow.

Given X-ray crystallography data of a protein molecule from the PDB, we build a van der Waal weighted alpha shape representation of that protein molecule by constructing cells around each atom center. Thus, to each protein we associate a set of persistence diagrams (one for each dimension). Using elementary physical principles, we identify certain structural features of molecules that are conjectured to impact compressibility. A simple parameter search through the persistence diagrams isolates these features and provides a robust measure which exhibits remarkable linear correlation with experimentally computed protein compressibility. Here is a preprint.