Site menu:

New Orleans

OPT2023

We welcome you to participate in the 15th International OPT Workshop on Optimization for Machine Learning, to be held as a part of the NeurIPS 2023 conference. This year we particularly encourage (but not limit) submissions inspired by the spirit of "Optimization in the Wild".


We are looking forward to an exciting OPT!


Schedule

Locaction: New Orleans Convention Center.

The schedule is also available on the NeurIPS virtual platform.

Time Speaker Title


Session 1 (Moderator: Courtney Paquette)

9:00am-9:05am Organizers Welcome Remarks
9:00am-9:30amYair Carmon (Tel-Aviv University) DoG is SGD’s best friend: toward tuning-free stochastic optimization[abstract]
9:30am-10:00amContributed talks
  • Bruno Loureiro: Escaping mediocrity: how two-layer networks learn hard generalized linear models
  • Daniil Vankov: Last Iterate Convergence of Popov Method for Non-monotone Stochastic Variational Inequalities
[papers]
10:00am-11:00amPoster Session 1 [posters]


Session 2 (Moderator: Aaron Sidford)

11:00am-11:30amContributed talks
  • Guy Kornowski: An Algorithm with Optimal Dimension-Dependence for Zero-Order Nonsmooth Nonconvex Stochastic Optimization
  • Michael Lu: Practical Principled Policy Optimization for Finite MDPs
[papers]
11:30am-12:00pmDmitry Drusvyatskiy (University of Washington) Aiming towards the minimizers: fast convergence of SGD for overparameterized problems[abstract]
12:00pm-02:00pmLunch


Session 3 (Moderator: Sebastian Stich)

02:00pm-02:30pmVirginia Smith (Carnegie Mellon University) Evaluating Large-Scale Learning Systems[abstract]
02:30pm-03:00pmContributed talks
  • Naren Sarayu Manoj: Dueling Optimization with a Monotone Adversary
  • Georgy Noarov: High-Dimensional Prediction for Sequential Decision Making
[papers]
03:00pm-04:00pmPoster Session 2 [posters]


Session 4 (Moderator: Courtney Paquette)

04:00pm-04:30pmAshwin Pananjady (Georgia Tech) Sharply predicting the behavior of complex iterative algorithms with random data[abstract]
04:30pm-05:00pmJason Lee (Princeton University) Provable Feature Learning in Gradient Descent[abstract]
05:00pm-05:01pmCristóbal Guzmán Closing Remarks