Research

My primary research interest is in algebraic combinatorics. I like thinking about polynomials--my favorite approaches to studying them involve combining the insights offered by abstract structures with computational methods. I am currently working on understanding homological invariants of various ideals stable under large group actions, particularly those that arise in Schubert calculus.

Publications:

We study the Robinson-Schensted-Knuth correspondence (RSK) as the transition operator between the monomial and bitableau bases of polynomial functions on the space of mxn complex-valued matrices. Our results characterize the diagonalizability of RSK_{m, n, d} (the restriction of RSK to degree-d homogeneous polynomials) and give formulas for its trace and determinant more efficient than the naïve ones. We also give several conjectures and open problems for future research.

We describe a "filtered" generalization of RSK and use it to prove a combinatorial rule for the multiplicities of irreducible representations in the coordinate rings of a class of matrix varieties we call "bicrystalline". We show that matrix Schubert varieties and their unions are bicrystalline. Our rule generalizes the classical Cauchy identity, Littlewood-Richardson rule, and the Knutson-Miller Hilbert series formula for matrix Schubert varieties. Slides and video from a (virtual) talk I gave at the Michigan State University Combinatorics and Graph Theory seminar are available.

We prove a sharp lower bound on the number of terms in an element of the reduced Gröbner basis of a Schubert determinantal ideal. We give three applications, including a pattern-avoidance characterization of toric matrix Schubert varieties that complements work of Escobar-Mészáros.

We describe a general algorithm for generating sets of "hieroglyphs" that compute the (multi)degree of a given variety. 

We prove that the Hilbert functions of Schubert determinantal ideals are polynomial using a simple bound on the Castelnuovo-Mumford regularity of these ideals.

We characterize a family of ideals called nearly complete ideals by representing them as certain hypergraphs. We use this characterization to describe some features of their Betti numbers.