Abstract:Models of stochastic processes are widely used in almost all fields of science. Theory validation, parameter estimation, and prediction all require model calibration and statistical inference using data. However, data are almost always incomplete observations of reality. This leads to a great challenge for statistical inference because the likelihood function will be intractable for almost all partially observed stochastic processes. This renders many statistical methods, especially within a Bayesian framework, impossible to implement. Therefore, computationally expensive likelihood-free approaches are applied that replace likelihood evaluations with realisations of the model and observation process. For accurate inference, however, likelihood-free techniques may require millions of expensive stochastic simulations. To address this challenge, we develop a new method based on recent advances in multilevel and multifidelity. Our approach combines the multilevel Monte Carlo telescoping summation, applied to a sequence of approximate Bayesian posterior targets, with a multifidelity rejection sampler to minimise the number of computationally expensive exact simulations required for accurate inference. We present the derivation of our new algorithm for likelihood-free Bayesian inference, discuss practical implementation details, and demonstrate substantial performance improvements. Using examples from systems biology, we demonstrate improvements of more than two orders of magnitude over standard rejection sampling techniques. Our approach is generally applicable to accelerate other sampling schemes, such as sequential Monte Carlo, to enable feasible Bayesian analysis for realistic practical applications in physics, chemistry, biology, epidemiology, ecology and economics.

Bayesian leave-one-out cross-validation for large data

Efficient leave-one-out cross-validation for Bayesian non-factorized normal and Student-t models

Practical Bayesian model evaluation using leave-one-out cross-validation and WAIC

Bayesian leave-one-out cross-validation approximations for Gaussian latent variable models

Approximate leave-future-out cross-validation for Bayesian time series models

Approximate Leave-one-out Cross Validation for Regression with $\ell_1$ Regularizers (extended version)

Theoretical Analysis of Leave-one-out Cross Validation for Non-differentiable Penalties under High-dimensional Settings

Efficient Selection Between Hierarchical Cognitive Models: Cross-validation With Variational Bayes

Bayesian nonparametric cross-study validation of prediction methods

Bayesian cross-validation by parallel Markov chain Monte Carlo

Modify Leave-One-Out Cross Validation by Moving Validation Samples Around Random Normal Distributions: Move-One-Away Cross Validation

Gradient-flow adaptive importance sampling for Bayesian leave one out cross-validation with application to sigmoidal classification models

Outlier Detection in Near-Infrared Spectroscopic Analysis by Using Monte Carlo Cross-Validation

Large Language Models to Enhance Bayesian Optimization

posteriordb: Testing, Benchmarking and Developing Bayesian Inference Algorithms

BLoB: Bayesian Low-Rank Adaptation by Backpropagation for Large Language Models

Efficient strategies for leave-one-out cross validation for genomic best linear unbiased prediction

Approximate Laplace approximations for scalable model selection

Bayesian Restricted Likelihood Methods: Conditioning on Insufficient Statistics in Bayesian Regression

Scalable Approximations of Marginal Posteriors in Variable Selection

Multifidelity multilevel Monte Carlo to accelerate approximate Bayesian parameter inference for partially observed stochastic processes