Please enable JavaScript.

Coggle requires JavaScript to display documents.

Conjugate Priors (Informative Prior (Type (Power Prior : make use of…

- - - - Power Prior = Likelihood (data) × Uninformative Conjugate Prior
- - - - Type
        
        Uniform Prior for Normal likelihood - f(μ) = 1
        
        f(σ^2) = 1/σ^2 is used for σ^2
  - - - Type
        
        Jeffreys Prior - Beta(0.5, 0.5)
        
        Given distribution f(x|θ), the Jeffreys prior is √(-E(∂^2/(∂θ^2 ) log⁡(f(x│θ)))
        
        Reference Priors : can be Uniform or Jeffreys Prior, depends on the likelihood.
        
        Uniform Prior - Beta(1, 1), f(θ) = 1