AMUSE

Date: February 22, 2023

Time: 6:00PM - 7:00PM

Location: BLOC 306

Speaker: Stephan Wojtowytsch, Texas A&M University

Title: Neural networks and how to train them

Abstract: The most time-consuming part of deep learning is the training of neural networks. We give a gentle introduction to data science, machine learning and deep learning. Our focus will be on training algorithms modeled on a hiker lost in the fog (gradient descent) and a heavy ball rolling down a mountainside (gradient descent with momentum) and their advantages.