Search results for "Importance sampling"

showing 4 items of 24 documents

Cross Correlations in Scaling Analyses of Phase Transitions

2008

Thermal or finite-size scaling analyses of importance sampling Monte Carlo time series in the vicinity of phase transition points often combine different estimates for the same quantity, such as a critical exponent, with the intent to reduce statistical fluctuations. We point out that the origin of such estimates in the same time series results in often pronounced cross-correlations which are usually ignored even in high-precision studies, generically leading to significant underestimation of statistical fluctuations. We suggest to use a simple extension of the conventional analysis taking correlation effects into account, which leads to improved estimators with often substantially reduced …

Statistical Mechanics (cond-mat.stat-mech)Monte Carlo methodFOS: Physical sciencesGeneral Physics and AstronomyStatistical fluctuationsDynamic Monte Carlo methodMonte Carlo method in statistical physicsStatistical physicsCritical exponentScalingCondensed Matter - Statistical MechanicsImportance samplingMonte Carlo molecular modelingMathematicsPhysical Review Letters
researchProduct

Importance sampling correction versus standard averages of reversible MCMCs in terms of the asymptotic variance

2017

We establish an ordering criterion for the asymptotic variances of two consistent Markov chain Monte Carlo (MCMC) estimators: an importance sampling (IS) estimator, based on an approximate reversible chain and subsequent IS weighting, and a standard MCMC estimator, based on an exact reversible chain. Essentially, we relax the criterion of the Peskun type covariance ordering by considering two different invariant probabilities, and obtain, in place of a strict ordering of asymptotic variances, a bound of the asymptotic variance of IS by that of the direct MCMC. Simple examples show that IS can have arbitrarily better or worse asymptotic variance than Metropolis-Hastings and delayed-acceptanc…

Statistics and ProbabilityFOS: Computer and information sciencesdelayed-acceptanceMarkovin ketjut01 natural sciencesStatistics - Computationasymptotic variance010104 statistics & probabilitysymbols.namesake60J22 65C05unbiased estimatorFOS: MathematicsApplied mathematics0101 mathematicsComputation (stat.CO)stokastiset prosessitestimointiMathematicsnumeeriset menetelmätpseudo-marginal algorithmApplied Mathematics010102 general mathematicsProbability (math.PR)EstimatorMarkov chain Monte CarloCovarianceInfimum and supremumWeightingMarkov chain Monte CarloMonte Carlo -menetelmätDelta methodimportance samplingModeling and SimulationBounded functionsymbolsImportance samplingMathematics - Probability
researchProduct

Importance sampling type estimators based on approximate marginal Markov chain Monte Carlo

2020

We consider importance sampling (IS) type weighted estimators based on Markov chain Monte Carlo (MCMC) targeting an approximate marginal of the target distribution. In the context of Bayesian latent variable models, the MCMC typically operates on the hyperparameters, and the subsequent weighting may be based on IS or sequential Monte Carlo (SMC), but allows for multilevel techniques as well. The IS approach provides a natural alternative to delayed acceptance (DA) pseudo-marginal/particle MCMC, and has many advantages over DA, including a straightforward parallelisation and additional flexibility in MCMC implementation. We detail minimal conditions which ensure strong consistency of the sug…

Statistics and ProbabilityHyperparameter05 social sciencesBayesian probabilityStrong consistencyEstimatorContext (language use)Markov chain Monte Carlo01 natural sciencesStatistics::Computation010104 statistics & probabilitysymbols.namesake0502 economics and businesssymbols0101 mathematicsStatistics Probability and UncertaintyParticle filterAlgorithmImportance sampling050205 econometrics MathematicsScandinavian Journal of Statistics
researchProduct

Event generation and statistical sampling for physics with deep generative models and a density information buffer

2021

Simulating nature and in particular processes in particle physics require expensive computations and sometimes would take much longer than scientists can afford. Here, we explore ways to a solution for this problem by investigating recent advances in generative modeling and present a study for the generation of events from a physical process with deep generative models. The simulation of physical processes requires not only the production of physical events, but to also ensure that these events occur with the correct frequencies. We investigate the feasibility of learning the event generation and the frequency of occurrence with several generative machine learning models to produce events l…

Test data generationScienceMonte Carlo methodGeneral Physics and AstronomyFOS: Physical sciences01 natural sciencesCharacterization and analytical techniquesGeneral Biochemistry Genetics and Molecular BiologyArticleHigh Energy Physics - ExperimentHigh Energy Physics - Experiment (hep-ex)High Energy Physics - Phenomenology (hep-ph)0103 physical sciencesInformation theory and computationHigh Energy Physics010306 general physicsMultidisciplinary010308 nuclear & particles physicsEvent (computing)QStatisticsData ScienceSampling (statistics)General ChemistryDensity estimationAutoencoderHigh Energy Physics - PhenomenologyPhysics - Data Analysis Statistics and ProbabilityExperimental High Energy PhysicsAnomaly detectionAlgorithmImportance samplingData Analysis Statistics and Probability (physics.data-an)
researchProduct