Dynamic Denoising for AMP Applied to Sparse Regression Inner Codes with Outer Codes

Explore a 36-minute lecture on dynamic denoising for approximate message passing (AMP) applied to sparse regression inner codes with outer codes. Delve into the successful application of AMP to sparse regression codes and learn how outer codes improve their finite-length performance. Discover the new paradigm of dynamic denoising, which integrates outer code structure into state estimation for enhanced performance. Examine potential extensions of this methodology, including state evolution approximations for outer code design, applications to multi-user scenarios, and adaptations for multiple measurement vectors. Join Jean-Francois Chamberland from Texas A&M University as he presents these concepts in the context of application-driven coding theory at the Simons Institute.

Dynamic Denoising For Amp Applied To Sparse Regression Inner Codes With Outer Codes