Loading Events

« All Events

  • This event has passed.

Laurent El Ghaoui, UC Berkeley: Implicit Deep Learning

March 10 @ 11:00 am - 12:30 pm

ABSTRACT: We define a new class of “implicit” deep learning prediction rules that generalize the recursive rules of feedforward neural networks. These models are based on the solution of a fixed-point equation involving a single vector of hidden features, which is thus only implicitly defined. The new framework greatly simplifies the notation of deep learning, and opens up many new possibilities, in terms of novel architectures and algorithms, robustness analysis, adversarial attacks, and design, interpretability, sparsity, and network  architecture optimization.

Working paper

Slides

Details

Date:
March 10
Time:
11:00 am - 12:30 pm
Event Category:

Venue

Online via Zoom