INTRODUCTION TO DEEP NEURAL NETWORKS.

Catalog description

Pre-requisites: EENG 383 and EENG 388 (or MATH 380 ); and EENG 255 (or CSCD 240 , or any high-level programming language such as C/C++, Java, Python etc.). Corequisite: EENG 427L . Provides an introduction to deep neural networks (DNNs) such as CNNs, RNNs, ResNets, GANs, etc. Those DNNs are built up from a basic multi-layer perceptron. The learning algorithm using backpropagation is introduced and built up to advanced learning algorithms such as SGD, Adam etc. In addition, several design issues in DNNs such as overfitting/underfitting, vanishing and exploding gradient problems etc. are explained in the context of optimization for DNNs. Companion course to EENG 427L .