Gravitational Wave Parameter Estimation with Compressed Likelihood Evaluations

Explore gravitational wave parameter estimation techniques in this 42-minute lecture by Scott Field from the University of Massachusetts Dartmouth. Delve into the challenges of Bayesian parameter estimation in gravitational wave astronomy and learn about Reduced Order Quadratures (ROQs) as a solution for fast and accurate likelihood evaluations. Discover the key algorithms and software needed to build ROQ rules, and gain insights into their applications for various gravitational wave models. Examine the computational benefits of ROQs, including their exponential convergence and potential to significantly accelerate inference runs. Investigate new challenges and opportunities for ROQs in upcoming detectors, and understand their importance in advancing gravitational wave research.

Intro
Gravitational wave datasets
Bayesian inference of GW datasets
Likelhood computations are too slow
Parameter estimation challenges
Approaches to faster PE (non-exhausthe list)
Reduced order quadratures (ROQS) in use
Outline
Numerical integration (quadrature)
Do I need a low-order quadrature rule for noisy data?
Probler Formulation
Step 1: Compressing the model
Best approximation space X
Example basis generation
Waveform compression application (ex: 1.2040)
Summary of step 1
Where are the good points for integrating in X.?
Empirical interpolation method
Example: Points for polynomial interpolation integratior
The ROQ approximation
Using ROQ
Building ROQ
Startup a signal has been detected!
How much faster?
Accelerating tests of GR
BNS events with third generation observatories