Assigning a value to a power likelihood in a general Bayesian model

Abstract

Bayesian approaches to data analysis and machine learning are widespread and popular as they provide intuitive yet rigorous axioms for learning from data; see Bernardo and Smith (2004) and Bishop (2006). However, this rigour comes with a caveat that the Bayesian model is a precise reflection of Nature. There has been a recent trend to address potential model misspecification by raising the likelihood function to a power, primarily for robustness reasons, though not exclusively. In this paper we provide a coherent specification of the power parameter once the Bayesian model has been specified in the absence of a perfect model.

References in corpus (1)

• From $\epsilon$-entropy to KL-entropy: Analysis of minimum information complexity density estimation

Cited by in corpus (10)

• Universal Inference
• Gibbs posterior concentration rates under sub-exponential type losses
• Bayesian model comparison with the Hyvärinen score: computation and consistency
• Learning Attribute Patterns in High-Dimensional Structured Latent Attribute Models
• Finite mixture models do not reliably learn the number of components
• Empirical priors and coverage of posterior credible sets in a sparse normal mean model
• Generalized Bayesian Likelihood-Free Inference
• Scalable high-dimensional Bayesian varying coefficient models with unknown within-subject covariance
• General Robust Bayes Pseudo-Posterior: Exponential Convergence results with Applications
• Towards a Theoretical Understanding of the Robustness of Variational Autoencoders