A Scalable Gradient-Free Method for Bayesian Experimental Design with Implicit Models

03/14/2021
by   Jiaxin Zhang, et al.
0

Bayesian experimental design (BED) is to answer the question that how to choose designs that maximize the information gathering. For implicit models, where the likelihood is intractable but sampling is possible, conventional BED methods have difficulties in efficiently estimating the posterior distribution and maximizing the mutual information (MI) between data and parameters. Recent work proposed the use of gradient ascent to maximize a lower bound on MI to deal with these issues. However, the approach requires a sampling path to compute the pathwise gradient of the MI lower bound with respect to the design variables, and such a pathwise gradient is usually inaccessible for implicit models. In this paper, we propose a novel approach that leverages recent advances in stochastic approximate gradient ascent incorporated with a smoothed variational MI estimator for efficient and robust BED. Without the necessity of pathwise gradients, our approach allows the design process to be achieved through a unified procedure with an approximate gradient for implicit models. Several experiments show that our approach outperforms baseline methods, and significantly improves the scalability of BED in high-dimensional problems.

READ FULL TEXT

page 1

page 2

page 3

page 4

research
03/14/2021

A Hybrid Gradient Method to Designing Bayesian Experiments for Implicit Models

Bayesian experimental design (BED) aims at designing an experiment to ma...
research
02/19/2020

Bayesian Experimental Design for Implicit Models by Mutual Information Neural Estimation

Implicit stochastic models, where the data-generation distribution is in...
research
10/23/2018

Efficient Bayesian Experimental Design for Implicit Models

Bayesian experimental design involves the optimal allocation of resource...
research
05/10/2021

Gradient-based Bayesian Experimental Design for Implicit Models using Mutual Information Lower Bounds

We introduce a framework for Bayesian experimental design (BED) with imp...
research
10/20/2020

Neural Approximate Sufficient Statistics for Implicit Models

We consider the fundamental problem of how to automatically construct su...
research
10/26/2021

Online Variational Filtering and Parameter Learning

We present a variational method for online state estimation and paramete...
research
06/06/2018

Randomized Value Functions via Multiplicative Normalizing Flows

Randomized value functions offer a promising approach towards the challe...

Please sign up or login with your details

Forgot password? Click here to reset