Loss-Based Variational Bayes Prediction

2021/04/29 by Frazier, David T., Loaiza-Maya, Ruben, Martin, Gael M. +1 · 2 citations
#Applications (stat.AP) #Econometrics (econ.EM) #FOS: Computer and information sciences #FOS: Economics and business #Methodology (stat.ME)

paper · doi:10.48550/arxiv.2104.14054

Abstract

We propose a new approach to Bayesian prediction that caters for models with a large number of parameters and is robust to model misspecification. Given a class of high-dimensional (but parametric) predictive models, this new approach constructs a posterior predictive using a variational approximation to a generalized posterior that is directly focused on predictive accuracy. The theoretical behavior of the new prediction approach is analyzed and a form of optimality demonstrated. Applications to both simulated and empirical data using high-dimensional Bayesian neural network and autoregressive mixture models demonstrate that the approach provides more accurate results than various alternatives, including misspecified likelihood-based predictions.

Cited by

• Generalized Bayesian Inference for Dynamic Random Dot Product Graphs
• Generative Regression with IQ-BART

Related

• Variational Prediction
• Predictive variational inference: Learn the predictively optimal posterior distribution
• Learning under Model Misspecification: Applications to Variational and Ensemble methods
• Variational Bayes Neural Network: Posterior Consistency, Classification Accuracy and Computational Challenges
• Variational Bayes under Model Misspecification