Abstract

Deep learning has become a powerful tool for cutting-edge machine learning and artificial intelligence applications such as data classification and computer vision. Deep learning uses neural networks comprising many trainable parameters, which gives rise to significant computational complexity, particularly during their training phase. Such complexity is often prohibitively large given the available time, financial and computational resources, and thereby restricts the study and usage of deep learning systems.

Date

2020

Authors

Sourya Dey

Institution

University of Southern California