ICML 2018

Scalable Gaussian Processes with Grid-Structured Eigenfunctions (GP-GRIEF)

ICML 2018 treforevans/gp_grief

We introduce a kernel approximation strategy that enables computation of the Gaussian process log marginal likelihood and all hyperparameter derivatives in $\mathcal{O}(p)$ time.

BAYESIAN INFERENCE

On the Power of Over-parametrization in Neural Networks with Quadratic Activation

ICML 2018 Clumsyndicate/One_layer_analysis_network

We provide new theoretical insights on why over-parametrization is effective in learning neural networks.

Comparing Dynamics: Deep Neural Networks versus Glassy Systems

ICML 2018 mbaityje/DEEP-GLASS

We analyze numerically the training dynamics of deep neural networks (DNN) by using methods developed in statistical physics of glassy systems.