ML Reviews

bayesian-parameter-estimation

Obtain an estimate for parameters θ\theta given data DD, by integrating over the full parameter space and evaluating p(θD)p(\theta \vert D) (i.e. the posterior).

With p(θD)p(\theta \vert D), you can get various point estimates:

  1. The expectation value, θp(θD)dθ\int \theta p(\theta \vert D) d\theta
  2. The [[maximum-a-posteriori]] estimate