Approximate Bayesian computation (ABC) gives exact results under the assumption of model error

Abstract

Approximate Bayesian computation (ABC) or likelihood-free inference algorithms are used to find approximations to posterior distributions without making explicit use of the likelihood function, depending instead on simulation of sample data sets from the model. In this paper we show that under the assumption of the existence of a uniform additive model error term, ABC algorithms give exact results when sufficient summaries are used. This interpretation allows the approximation made in many previous application papers to be understood, and should guide the choice of metric and tolerance in future work. ABC algorithms can be generalized by replacing the 0-1 cut-off with an acceptance probability that varies with the distance of the simulated data from the observed data. The acceptance density gives the distribution of the error term, enabling the uniform error usually used to be replaced by a general distribution. This generalization can also be applied to approximate Markov chain Monte Carlo algorithms. In light of this work, ABC algorithms can be seen as calibration techniques for implicit stochastic models, inferring parameter values in light of the computer model, data, prior beliefs about the parameter values, and any measurement or model errors.

33 pages, 1 figure, to appear in Statistical Applications in Genetics and Molecular Biology 2013

References in corpus (7)

• Approximate Bayesian computation scheme for parameter inference and model selection in dynamical systems
• Inferring population history with DIYABC: a user-friendly approach to Approximate Bayesian Computation
• Adaptive approximate Bayesian computation
• Non-linear regression models for Approximate Bayesian Computation
• MCMC for doubly-intractable distributions
• HIV with contact-tracing: a case study in Approximate Bayesian Computation
• Semi-automatic selection of summary statistics for ABC model choice

Cited by in corpus (50)

• Non-linear regression models for Approximate Bayesian Computation
• Simulation and inference algorithms for stochastic biochemical reaction networks: from basic concepts to state-of-the-art
• Bayesian design of synthetic biological systems
• Bayesian uncertainty analysis for complex systems biology models: emulation, global parameter searches and evaluation of gene functions
• Likelihood-free Markov chain Monte Carlo
• A Unified Bayesian Inference Framework for Generalized Linear Models
• A Simulated Annealing Approach to Approximate Bayes Computations
• How can a glacial inception be predicted?
• Accelerating ABC methods using Gaussian processes
• Considering discrepancy when calibrating a mechanistic electrophysiology model
• Inference for SDE models via Approximate Bayesian Computation
• Approximate Bayesian Image Interpretation using Generative Probabilistic Graphics Programs
• Constructing Summary Statistics for Approximate Bayesian Computation: Semi-automatic ABC
• Efficient acquisition rules for model-based approximate Bayesian computation
• Overview of Approximate Bayesian Computation
• Black-box Bayesian inference for economic agent-based models
• Bayesian model comparison with un-normalised likelihoods
• Bayesian Restricted Likelihood Methods: Conditioning on Insufficient Statistics in Bayesian Regression
• A practical guide to pseudo-marginal methods for computational inference in systems biology
• Asymptotically exact data augmentation: models, properties and algorithms
• Approximating Bayes in the 21st Century
• Bayesian computation: a perspective on the current state, and sampling backwards and forwards
• Conditional Random Matrix Ensembles and the Stability of Dynamical Systems
• Notes to Robert et al.: Model criticism informs model choice and model comparison
• Batch simulations and uncertainty quantification in Gaussian process surrogate approximate Bayesian computation
• Generalized Posteriors in Approximate Bayesian Computation
• Improvements to Inference Compilation for Probabilistic Programming in Large-Scale Scientific Simulators
• Automatic Variational ABC
• Fast Approximate Bayesian Computation for discretely observed Markov models using a factorised posterior distribution
• Deviance Information Criteria for Model Selection in Approximate Bayesian Computation
• Approximate Bayesian Computation via Classification
• Parameter Estimation in Hidden Markov Models with Intractable Likelihoods Using Sequential Monte Carlo
• AABC: approximate approximate Bayesian computation when simulating a large number of data sets is computationally infeasible
• Statistical inference for dynamical systems: a review
• A Simple Approach to Maximum Intractable Likelihood Estimation
• Approximate Bayesian Computation: a nonparametric perspective
• Simple techniques for likelihood analysis of univariate and multivariate stable distributions: with extensions to multivariate stochastic volatility and dynamic factor models
• Approximate Bayesian algorithms for multiple target tracking with binary sensors
• Reliable ABC model choice via random forests
• Approximate Bayesian Computation for a Class of Time Series Models
• Mutation and selection in bacteria: modelling and calibration
• A review of Approximate Bayesian Computation methods via density estimation: inference for simulator-models
• POPE: Post Optimization Posterior Evaluation of Likelihood Free Models
• Hierarchical surrogate-based Approximate Bayesian Computation for an electric motor test bench
• Classification of Computer Models with Labelled Outputs
• Using Approximate Bayesian Computation by Subset Simulation for Efficient Posterior Assessment of Dynamic State-Space Model Classes
• Detecting conflicting summary statistics in likelihood-free inference
• Modelling Numerical Systems with Two Distinct Labelled Output Classes
• Understanding Hormonal Crosstalk in Arabidopsis Root Development via Emulation and History Matching
• Embedded Model Form Uncertainty Quantification with Measurement Noise for Bayesian Model Calibration