Oct 16, 2018 · We give a new algorithm for learning a two-layer neural network under a general class of input distributions.

We give a new algorithm for learning a two-layer neural network under a general class of input distributions. Assuming there is a ground-truth two-layer ...

When the input distribution is symmetric, we give the first algorithm that can learn a two-layer neural network. Our algorithm is based on the method-of-moments ...

A new algorithm for learning a two-layer neural network under a general class of input distributions based on the method-of-moments framework and extends ...

Oct 17, 2018 · We give a new algorithm for learning a two-layer neural network under a general class of input distributions.

Nov 20, 2019 · That is to say, if π(⋅) is a particular permutation, then π(X) should output π(Y). So each input should be treated symmetrically in some sense.

Sep 22, 2019 · How to exploit logical symmetry of input neurons to train a supervised neural network? ... Mostly as an exercise, I am training a neural network ...

We give a new algorithm for learning a two-layer neural network under a very general class of input distributions. Assuming there is a ground-truth two-layer ...

Apr 26, 2022 · Bibliographic details on Learning Two-layer Neural Networks with Symmetric Inputs.