Mini-batch SGD with momentum is a fundamental algorithm for learning lar...
A memory efficient approach to ensembling neural networks is to share mo...
Performance of optimization on quadratic problems sensitively depends on...
Current theoretical results on optimization trajectories of neural netwo...
We call a finite family of activation functions superexpressive if any
m...
Recent research shows that sublevel sets of the loss surfaces of
overpar...
In this paper, we propose a new nonlinear detector with improved interfe...
We explore the phase diagram of approximation rates for deep neural netw...
We derive a nonlinear integro-differential transport equation describing...
We describe generalizations of the universal approximation theorem for n...
We prove that deep ReLU neural networks with conventional fully-connecte...
We consider approximations of 1D Lipschitz functions by deep ReLU networ...
We introduce a library of geometric voxel features for CAD surface
recog...
We study expressive power of shallow and deep neural networks with piece...
We describe GTApprox - a new tool for medium-scale surrogate modeling in...