BIMSA > Seminar on Control Theory and Nonlinear Filtering Deep Learning without Global Optimization by Random Fourier Neural Networks
Deep Learning without Global Optimization by Random Fourier Neural Networks
Organizer
Shing Toung Yau
Speaker
Jia Yi Kang
Time
Wednesday, July 24, 2024 2:30 PM - 3:00 PM
Venue
Online
Abstract
We introduce a new training algorithm for variety of deep neural networks that utilize random complex exponential activation functions. Our approach employs a Markov Chain Monte Carlo sampling procedure to iteratively train network layers, avoiding global and gradient-based optimization while maintaining error control. It consistently attains the theoretical approximation rate for residual networks with complex exponential activation functions, determined by network complexity. Additionally, it enables efficient learning of multiscale and high-frequency features, producing interpretable parameter distributions. Despite using sinusoidal basis functions, we do not observe Gibbs phenomena in approximating discontinuous target functions.