Mathematical Foundations of Learning from Data and Signals (Math-FLDS)

This is a biweekly lunch reading group for graduate students/postdocs and faculty at USC to learn about some recent advances in mathematical data analysis. In Fall of 2016 this lunch series is from 12:00-2:00 PM in EEB 248.

Week 1 (October 13, 2016): Introductory meeting

Week 2 (October 27, 2016): Aamir Anis, Deep Feedforward Networks and Regularization for Deep Learning (Chapters 6 and 7 from Deep Learning)

Week 3 (November 3, 2016): Akshay Gadde, Optimization for Training Deep Models(Chapter 8 from Deep Learning)

Week 4 (November 10, 2016): Convolutional Networks and Sequence Modeling: Recurrent and Recursive Nets (Chapters 9 and 10 from Deep Learning)

Week 5 (December 1, 2016): Antonio Ortega Scattering transforms

Week 6 (December 8, 2016): Mahdi Soltanolkotabi On the dynamics of training of neural networks