Deep Neural Networks in a Mathematical Framework

This SpringerBrief describes how to build a rigorous end-to-end mathematical framework for deep neural networks. In particular, the authors derive gradient descent algorithms in a unified way for several neural network structures, including multilayer perceptrons, convolutional neural networks, deep autoencoders and recurrent neural networks.

Acest SpringerBrief descrie cum sa construim un cadru matematic riguros de la capat la capat pentru retelele neuronale profunde. In special, autorii deriva algoritmi de coborare in gradient intr-un mod unificat pentru mai multe structuri de retele neuronale, incluzand perceptroni multistrat, retele neuronale convolutionale, autoencoderi adanci si retele neuronale recurente.