Title: Deep Learning: An Introduction from perspective of applied mathematics
Department of Mathematics, National University of Singapore
Date:2018/12/10, Monday, 10:00-11:00
Abstract: The recent rapid progress in data sciences and artificial intelligence largely relies on the advances of deep learning techniques, together with the availability of large datasets and specific hardware (GPU). This talk targets general audience who are keen to knowing about deep learning by providing a brief introduction to the basic ideas that underlie deep learning from an applied mathematics perspective. In particularly, we will focus on mathematical interpretation of the core concepts of deep learning, notably in calculus, approximation theory, optimization and harmonic analysis. Three fundamental questions will be touched in this talk: what is a deep neural network? how it is different from other modelling techniques network trained? What challenges we are facing when training a deep neural network.