Campuses:

Algorithms

Friday, November 9, 2018 - 11:10am - 11:40am
Edward McFowland (University of Minnesota, Twin Cities)
In the recent literature on estimating heterogeneous treatment effects, each proposed method makes its own set of restrictive assumptions about the intervention’s effects and which subpopulations to explicitly estimate. Moreover, the majority of the literature provides no mechanism to identify which subpopulations are the most affected–beyond manual inspection–and provides little guarantee on the correctness of the identified subpopulations.
Monday, August 10, 2015 - 10:50am - 11:10am
Friday, February 27, 2015 - 11:30am - 12:20pm
Nina Balcan (Carnegie-Mellon University)
Active learning is an important modern learning paradigm where the
algorithm itself can ask for labels of carefully chosen examples from a
large pool of unannotated data with the goal of minimizing human labeling
effort. In this talk, I will present a computationally efficient, noise
tolerant, and label efficient active learning algorithm for learning
linear separators under log-concave and nearly log-concave distributions.
Our technique exploits localization in several ways and can be thought of
Tuesday, February 24, 2015 - 2:00pm - 2:50pm
Jan Vondrak (IBM Research Division)
Much progress has been made on problems involving optimization of submodular functions under various constraints. However, the resulting algorithms, in particular the ones based on the multilinear relaxation, are often quite slow. In this talk, I will discuss some recent efforts on making these algorithms faster and more practical.
Tuesday, February 24, 2015 - 9:00am - 9:50am
Stefanie Jegelka (Massachusetts Institute of Technology)
Submodular functions capture a variety of discrete problems in machine learning, signal processing and computer vision. In these areas, practical algorithms are a major concern. Luckily, the submodular functions occurring in practice often have additional structure that can be exploited for practically efficient optimization algorithms.
Thursday, May 10, 2012 - 1:30pm - 2:30pm
Cynthia Rudin (Massachusetts Institute of Technology)
I will describe work on three areas related to crowd-based user-centered modeling:

1) Growing Lists:
We want to combining the knowledge of many people (experts) in order to create sets of things that go together, starting from a small seed. The experts have varying levels of expertise. This is the same problem that Google Sets was designed to solve. (With Ben Letham and Katherine Heller)

2) Sequential Event Prediction for Personalized Recommendations:
Thursday, May 28, 2015 - 4:30pm - 5:20pm
Kay Kirkpatrick (University of Illinois at Urbana-Champaign)
Near absolute zero, a gas of quantum particles can condense into an unusual state of matter, called Bose-Einstein condensation (BEC), that behaves like a giant quantum particle. The rigorous connection has recently been made between the physics of the microscopic many-body dynamics and the mathematics of the macroscopic model, the cubic nonlinear Schrodinger equation (NLS). I'll discuss progress with Gerard Ben Arous and Benjamin Schlein on a central limit theorem for the quantum many-body dynamics, a step towards large deviations for Bose-Einstein condensation.
Friday, May 30, 2014 - 11:00am - 12:30pm
Ben Recht (University of California, Berkeley)
Friday, September 14, 2012 - 4:30pm - 5:00pm
Qiang Du (The Pennsylvania State University)
To improve materials design, it is important to understand the influence of
microstructure on the physical properties of a material. Very often, it is the
nucleation process that dictates the microstructure. We present some recent
joint works with colleagues at Penn State on the computational studies of critical
nuclei morphology, growth and coarsening of microstructures as well as
stability of interacting particle systems.
Saturday, September 27, 2008 - 3:30pm - 4:30pm
Juan Meza (Lawrence Berkeley National Laboratory)
Combined with 9/26 abstract.

Pages

Subscribe to RSS - Algorithms