Machine Learning with Shallow Neural Networks

Abstract

Conventional machine learning often uses optimization and gradient-descent methods for learning parameterized models. Examples of such models include linear regression, support vector machines, logistic regression, dimensionality reduction, and matrix factorization....