Tractability And Overparametrization In Some Simple Models

preview_player
Показать описание
Andrea Montanari (Stanford University)
Algorithmic Advances for Statistical Inference with Combinatorial Structure
  1. Simons Institute
  2. Simons Institute
  3. theoretical computer science
  4. UC Berkeley
  5. Computer Science
  6. Theory of Computation
Рекомендации по теме
Tractability And Overparametrization 0:45:16

Tractability And Overparametrization In Some Simple Models

Geometry and Local 1:15:31

Geometry and Local Recovery of Global Minima of Two-layer Neural Networks at Overparamet

What is the 0:00:36

What is the meaning of the word TRACTABILITY?

From overparametrized neural 0:55:05

From overparametrized neural networks to harmonic regression

Linearized Neural Networks 0:01:47

Linearized Neural Networks

Jason Lee (USC)--Towards 0:42:51

Jason Lee (USC)--Towards a Foundation of Deep Learning: SGD, Overparametrization, and Generalization

The generalization error 0:53:39

The generalization error of overparametrized models - Andrea Montanari (Stanford) MAD+ 24 Jun 2020

Optimal Iterative Algorithms 0:59:11

Optimal Iterative Algorithms for Problems With Random Data (continued)

Adversarial Robustness 0:30:55

Adversarial Robustness

Santosh Vempala (Gatech) 0:43:59

Santosh Vempala (Gatech) -- Is There a Tractable (and Interesting) Theory of Nonconvex Optimization?

【EP11】Improving Robustness to 0:31:31

【EP11】Improving Robustness to Distribution Shifts: Methods and Benchmarks

Beyond Lazy Training 0:37:13

Beyond Lazy Training for Over-parameterized Tensor Decomposition

Suvrit Sra: Two 1:03:59

Suvrit Sra: Two Surprises When Optimization Meets Machine Learning

[ICML 2020] Empirical 0:15:02

[ICML 2020] Empirical Study: the Benefits of Overparameterization in Learning Latent Variable Models

Grokking: Generalization beyond 0:29:47

Grokking: Generalization beyond Overfitting on small algorithmic datasets (Paper Explained)

Expressive yet Tractable 0:12:15

Expressive yet Tractable Bayesian Deep Learning via Subnetwork Inference (AABI 2020)

Nonconvex Optimization for 0:34:41

Nonconvex Optimization for High-dimensional Learning: From ReLUs to Submodular Maximization

Fit Without Fear: 0:55:13

Fit Without Fear: An Over-Fitting Perspective on Modern Deep and Shallow Learning

Tradeoffs between Robustness 1:03:07

Tradeoffs between Robustness and Accuracy - Percy Liang

4. Recent Progress 1:19:14

4. Recent Progress In The Theory Of Deep Learning

Provable Robustness Beyond 0:47:56

Provable Robustness Beyond Bound Propagation

Toward a Causal 0:33:18

Toward a Causal Analysis of Generalization in Deep Learning - Behnam Neyshabur

Towards demystifying over-parameterization 0:43:47

Towards demystifying over-parameterization in deep (...) - Soltanolkotabi - Workshop 3 - CEB T1 2019

Minimum Complexity Interpolation 0:21:58

Minimum Complexity Interpolation in Random Features Models