Variational Models and Algorithms for GW Denoising and Reconstruction

Explore variational models and algorithms for gravitational wave (GW) denoising and reconstruction in this 40-minute conference talk by Alejandro Torres-Forné from the University of Valencia. Delve into the application of various methods to GW signals obtained from simulations under different noise conditions, comparing results from Gaussian to real noise. Examine how dictionary learning methods can be applied to classify different classes of signals and glitches. Learn about GW signal detection, data analysis steps, and signal denoising approaches, including TV methods, Rudin-Osher-Fatemi model, and Split-Bregman method. Investigate sparse representation of signals, The LASSO, and Dictionary Learning problem. Discover the process of searching for optimal regularization parameters and integrating with CWB. Explore CCSN mechanism extraction using LASSO and Dictionary Learning, as well as lip denoising via dictionary learning. Gain insights into the latest advancements in GW data analysis and reconstruction techniques presented at IPAM's Workshop IV: Big Data in Multi-Messenger Astrophysics.

Intro
GW signal detection
GW data analysis steps
Signal denoising approach
Introduction to TV methods
Rudin-Osher-Fatemi model
Split-Bregman method
Sparse representation of signals
The LASSO
Dictionary Learning problem
Search Optimal Regularization Parameter
Integration with CWB
Learning process
Dictionary learning results
CCSN mechanism extraction with LASSO
CCSN mechanism extraction with DL
lip denoising via dictionary learning
ummary and Conclusions