A Comparative Review of Dimension Reduction Methods in Approximate Bayesian Computation

Abstract

Approximate Bayesian computation (ABC) methods make use of comparisons between simulated and observed summary statistics to overcome the problem of computationally intractable likelihood functions. As the practical implementation of ABC requires computations based on vectors of summary statistics, rather than full data sets, a central question is how to derive low-dimensional summary statistics from the observed data with minimal loss of information. In this article we provide a comprehensive review and comparison of the performance of the principal methods of dimension reduction proposed in the ABC literature. The methods are split into three nonmutually exclusive classes consisting of best subset selection methods, projection techniques and regularization. In addition, we introduce two new methods of dimension reduction. The first is a best subset selection method based on Akaike and Bayesian information criteria, and the second uses ridge regression as a regularization procedure. We illustrate the performance of these dimension reduction techniques through the analysis of three challenging models and data sets.

Published in at http://dx.doi.org/10.1214/12-STS406 the Statistical Science (http://www.imstat.org/sts/) by the Institute of Mathematical Statistics (http://www.imstat.org)

References in corpus (4)

• Approximate Bayesian computation scheme for parameter inference and model selection in dynamical systems
• Expectation Propagation for approximate Bayesian inference
• A Comparative Review of Dimension Reduction Methods in Approximate Bayesian Computation
• Kernel Approximate Bayesian Computation for Population Genetic Inferences

Cited by in corpus (73)

• A Comparative Review of Dimension Reduction Methods in Approximate Bayesian Computation
• Massive optimal data compression and density estimation for scalable, likelihood-free inference in cosmology
• Bayesian Indirect Inference Using a Parametric Auxiliary Model
• Flexible statistical inference for mechanistic models of neural dynamics
• Approximate Bayesian Computation for Forward Modeling in Cosmology
• cosmoabc: Likelihood-free inference via Population Monte Carlo Approximate Bayesian Computation
• Model selection for hybrid dynamical systems via sparse regression
• A new model to predict weak-lensing peak counts II. Parameter constraint strategies
• Variational Bayes with Synthetic Likelihood
• Technical Note: Approximate Bayesian parameterization of a process-based tropical forest model
• Efficient acquisition rules for model-based approximate Bayesian computation
• ABC-CDE: Towards Approximate Bayesian Computation with Complex High-Dimensional Data and Limited Simulations
• Overview of Approximate Bayesian Computation
• Black-box Bayesian inference for economic agent-based models
• Summary Statistics in Approximate Bayesian Computation
• Neural Approximate Sufficient Statistics for Implicit Models
• Bayesian inference for stochastic differential equation mixed effects models of a tumor xenography study
• Rapid Bayesian inference for expensive stochastic models
• CoLFI: Cosmological Likelihood-free Inference with Neural Density Estimators
• Approximating Bayes in the 21st Century
• Indirect inference through prediction
• A Bayesian Optimization approach for calibrating large-scale activity-based transport models
• The impact of temporal sampling resolution on parameter inference for biological transport models
• Dynamic Likelihood-free Inference via Ratio Estimation (DIRE)
• DR-ABC: Approximate Bayesian Computation with Kernel-Based Distribution Regression
• Bayesian computation: a perspective on the current state, and sampling backwards and forwards
• Machine Learning Accelerated Likelihood-Free Event Reconstruction in Dark Matter Direct Detection
• Approximate maximum likelihood estimation using data-cloning ABC
• A Preferential Attachment Model for the Stellar Initial Mass Function
• Adaptive ABC model choice and geometric summary statistics for hidden Gibbs random fields
• Generalized Posteriors in Approximate Bayesian Computation
• Neural Density Estimation and Likelihood-free Inference
• Calibration of a bumble bee foraging model using Approximate Bayesian Computation
• Approximate Bayesian Computation by Modelling Summary Statistics in a Quasi-likelihood Framework
• On the Asymptotic Efficiency of Approximate Bayesian Computation Estimators
• Privacy-preserving Federated Bayesian Learning of a Generative Model for Imbalanced Classification of Clinical Data
• Approximate Bayesian computation and Bayes linear analysis: Towards high-dimensional ABC
• Distance-learning For Approximate Bayesian Computation To Model a Volcanic Eruption
• Sequentially guided MCMC proposals for synthetic likelihoods and correlated synthetic likelihoods
• Approximate methods for dynamic ecological models
• Vector operations for accelerating expensive Bayesian computations -- a tutorial guide
• Extending approximate Bayesian computation methods to high dimensions via a Gaussian copula model
• Regression approaches for Approximate Bayesian Computation
• An easy-to-use empirical likelihood ABC method
• Approximate Bayesian Computation via Classification
• Markov Chain Monte Carlo Methods, a survey with some frequent misunderstandings
• Data adaptation in HANDY economy-ideology model
• Comparing the morphology of molecular clouds without supervision
• Robust Optimisation Monte Carlo
• Assessment and adjustment of approximate inference algorithms using the law of total variance
• Multi-Statistic Approximate Bayesian Computation with Multi-Armed Bandits
• Reliable ABC model choice via random forests
• A review of Approximate Bayesian Computation methods via density estimation: inference for simulator-models
• Kernel-based Approximate Bayesian Inference for Exponential Family Random Graph Models
• Effective experience rating for large insurance portfolios via surrogate modeling
• A Guide to General-Purpose Approximate Bayesian Computation Software
• Deep inference of simulated strong lenses in ground-based surveys
• Recalibration: A post-processing method for approximate Bayesian computation
• Human-in-the-Loop Robot Planning with Non-Contextual Bandit Feedback
• On a Variational Approximation based Empirical Likelihood ABC Method
• GRAFFL: Gradient-free Federated Learning of a Bayesian Generative Model
• Likelihood-free approximate Gibbs sampling
• Approximate Bayesian Computation via Sufficient Dimension Reduction
• Likelihood free inference for Markov processes: a comparison
• Simultaneous adjustment of bias and coverage probabilities for confidence intervals
• A Return to Biased Nets: New Specifications and Approximate Bayesian Inference
• An adaptive mixture-population Monte Carlo method for likelihood-free inference
• Detecting conflicting summary statistics in likelihood-free inference
• Functional regression approximate Bayesian computation for Gaussian process density estimation
• A Parzen-based distance between probability measures as an alternative of summary statistics in Approximate Bayesian Computation
• Deep Fiducial Inference
• Hierarchical surrogate-based Approximate Bayesian Computation for an electric motor test bench
• Testing Inflationary Cosmology