Future students

Title: Tutorial on back-propagation and automatic differentiation

Abstract: In this presentation, I'll cover the basics of automatic differentiation (AD).

Title: Extending Thomassen’s Theorem to Two Faces

Abstract:

Let G be a planar graph and let L be a list-assignment for G in which there is a precolored edge on the outer face, every other vertex on the outer face has a list of size at least 3, and every other vertex in G has a list of size at least 5.

Title: Razborov's flag algebras: Ten years on

Abstract:

Many of the classical results in extremal combinatorics were obtained by ingenious application of elementary techniques, such as induction and Cauchy-Schwarz inequality.

Title: Primal-dual and Lagrangian relaxation techniques for k-median

Abstract: We will develop primal-dual algorithms to obtain constant-factor approximations for the uncapacitated facility location problem.

Title: New algorithms for maximum disjoint paths based on tree-likeness

Abstract:

We study the classical NP-hard problems of finding maximum-size subsets from given sets of k terminal pairs that can be routed via edge-disjoint paths (MaxEDP) or node-disjoint paths (MaxNDP) in a given graph.

Title: Basis Shape Loci and the Positive Grassmannian

Abstract: We study the set of k-dimensional planes in R^n admitting a basis of vectors with prescribed supports.

Title: Offline Assisted Group Key Exchange

Abstract:

This talk will focus on the problem of forward secrecy in group key exchange (GKE), where most of the participants remain offline until they wish to compute the key.

Title: Controllable graphs

Abstract: A graph is controllable if no eigenvector is orthogonal to the all-ones vector.

Title: Representable orientable matroids that are not real-representable

Abstract: In this talk we will have a brief introduction to oriented matroids and their relation to real-representability.

Title: Review of martingale theory, stochastic gradient descent, and adaptive line-line-search for stochastic optimization.

Abstract: With the rise of large data sets, practical algorithms for machine learning often use probability and statistics.