Approximate Bayesian Computation with Domain Expert in the Loop

Ayush Bharti, Louis Filstroff, Samuel Kaski
Proceedings of the 39th International Conference on Machine Learning, PMLR 162:1893-1905, 2022.

Abstract

Approximate Bayesian computation (ABC) is a popular likelihood-free inference method for models with intractable likelihood functions. As ABC methods usually rely on comparing summary statistics of observed and simulated data, the choice of the statistics is crucial. This choice involves a trade-off between loss of information and dimensionality reduction, and is often determined based on domain knowledge. However, handcrafting and selecting suitable statistics is a laborious task involving multiple trial-and-error steps. In this work, we introduce an active learning method for ABC statistics selection which reduces the domain expert’s work considerably. By involving the experts, we are able to handle misspecified models, unlike the existing dimension reduction methods. Moreover, empirical results show better posterior estimates than with existing methods, when the simulation budget is limited.

Cite this Paper

BibTeX
@InProceedings{pmlr-v162-bharti22a, title = {Approximate {B}ayesian Computation with Domain Expert in the Loop}, author = {Bharti, Ayush and Filstroff, Louis and Kaski, Samuel}, booktitle = {Proceedings of the 39th International Conference on Machine Learning}, pages = {1893--1905}, year = {2022}, editor = {Chaudhuri, Kamalika and Jegelka, Stefanie and Song, Le and Szepesvari, Csaba and Niu, Gang and Sabato, Sivan}, volume = {162}, series = {Proceedings of Machine Learning Research}, month = {17--23 Jul}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v162/bharti22a/bharti22a.pdf}, url = {https://proceedings.mlr.press/v162/bharti22a.html}, abstract = {Approximate Bayesian computation (ABC) is a popular likelihood-free inference method for models with intractable likelihood functions. As ABC methods usually rely on comparing summary statistics of observed and simulated data, the choice of the statistics is crucial. This choice involves a trade-off between loss of information and dimensionality reduction, and is often determined based on domain knowledge. However, handcrafting and selecting suitable statistics is a laborious task involving multiple trial-and-error steps. In this work, we introduce an active learning method for ABC statistics selection which reduces the domain expert’s work considerably. By involving the experts, we are able to handle misspecified models, unlike the existing dimension reduction methods. Moreover, empirical results show better posterior estimates than with existing methods, when the simulation budget is limited.} }
Endnote
%0 Conference Paper %T Approximate Bayesian Computation with Domain Expert in the Loop %A Ayush Bharti %A Louis Filstroff %A Samuel Kaski %B Proceedings of the 39th International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2022 %E Kamalika Chaudhuri %E Stefanie Jegelka %E Le Song %E Csaba Szepesvari %E Gang Niu %E Sivan Sabato %F pmlr-v162-bharti22a %I PMLR %P 1893--1905 %U https://proceedings.mlr.press/v162/bharti22a.html %V 162 %X Approximate Bayesian computation (ABC) is a popular likelihood-free inference method for models with intractable likelihood functions. As ABC methods usually rely on comparing summary statistics of observed and simulated data, the choice of the statistics is crucial. This choice involves a trade-off between loss of information and dimensionality reduction, and is often determined based on domain knowledge. However, handcrafting and selecting suitable statistics is a laborious task involving multiple trial-and-error steps. In this work, we introduce an active learning method for ABC statistics selection which reduces the domain expert’s work considerably. By involving the experts, we are able to handle misspecified models, unlike the existing dimension reduction methods. Moreover, empirical results show better posterior estimates than with existing methods, when the simulation budget is limited.
APA
Bharti, A., Filstroff, L. & Kaski, S.. (2022). Approximate Bayesian Computation with Domain Expert in the Loop. Proceedings of the 39th International Conference on Machine Learning, in Proceedings of Machine Learning Research 162:1893-1905 Available from https://proceedings.mlr.press/v162/bharti22a.html.

Related Material