| 08:30 - 08:58 | Coffee | |
| 08:58 – 09:00 | Opening remarks | |
| 09:00 – 09:30 |
Ahmet Alacaoglu |
Revisiting Inexact Fixed-Point Iterations for Min-Max Problems: Stochasticity and Structured Nonconvexity |
| 09:30 – 09:45 |
Yuan Gao |
On a result by Baillon, Bruck and Reich |
| 09:45 – 10:00 |
Ziyuan Wang |
Level proximal subdifferential, variational convexity, and pointwise Lipschitz smoothness |
| 10:00 – 10:30 |
Jiajin Li |
Spurious stationarity and hardness results for mirror descent |
| 10:30 - 11:00 | Break | |
| 11:00 – 11:30 |
Marco Caoduro |
A characterization of unimodular hypergraphs with disjoint hyperedges |
| 11:30 – 11:45 |
Yiwen Chen |
Characterizing the inscribability of polytopes using slack matrices |
| 11:45 – 12:00 |
Nicholas Richardson |
Density Separation with Tensor Factorization |
| 12:00 – 12:15 |
Richard Hoshino |
Cohort-Based Timetabling with Integer Linear Programming |
| 12:15 - 13:45 | Lunch | |
| 13:45 – 14:00 |
Alan Milligan |
Towards understanding the performance gap between Adam and SGD in deep learning |
| 14:00 – 14:15 |
Betty Shea |
Why Line Search when you can Plane Search? |
| 14:15 – 14:30 |
Begoña García Malaxechebarría |
The High Line: SGD dynamics of adaptive learning rate algorithms |
| 14:30 – 14:45 |
Brian Irwin |
EnKSGD: A Class of Preconditioned Black Box Optimization and Inversion Algorithms |
| 14:45 - 15:00 | Break | |
| 15:00 – 15:15 |
Tanmaya Karmarkar |
Computing the convex envelope of nonconvex piecewise bivariate linear-quadratic functions |
| 15:15 – 15:30 |
Max Neyra-Nesterenko |
Parameter-free and optimal restart schemes for first-order methods via approximate sharpness |
| 15:30 – 16:00 |
Robert Baraldi |
An Inexact Trust-Region Algorithm for Nonsmooth Nonconvex Regularized Problems |
| 16:00 – 16:30 |
Dima Drusvyatskiy |
The radius of statistical efficiency |
| 16:30 – | Adjournment | |
