MPoL-dev / MPoL

A flexible Python platform for Regularized Maximum Likelihood imaging
https://mpol-dev.github.io/MPoL/
MIT License
33 stars 11 forks source link

Audit "Intro to RML" document for proper usage of likelihood function notation #168

Closed iancze closed 1 year ago

iancze commented 1 year ago

After reading this helpful guide on Bayesian inference by Eadie et al. I realized I am one of those astronomers who writes likelihood functions inconsistently!

We should audit the Intro to RML document equations to make sure we're using consistent notation, especially for the likelihood function. Of their selected examples, I think I like $\mathcal{L}(\boldsymbol{\theta} ; \boldsymbol{y})$ the most, since it will lead us naturally to the "negative log likelihood loss" terminology that we use in a ML context.

Screenshot 2023-02-25 at 6 56 41 AM