Research on Control-variate Estimators of Variance Reduction
Fei Liu
Abstract
Fei Liu
Abstract
Variance Reduction(VR) is an important way to raise simulation estimation efficiency,and Control Variate(CV) estimators are widely used for variance reduction.Based on much works in developing and analyzing CV estimators,some further researches are given.Control variate estimation is studied where the control mean itself is unbiased estimated,and take both prior and auxiliary simulation replications into consideration.Analyzing control variate estimator is focus on where the control variance is known,and given a compound estimator depending on simulation replications and covariances.Then the efficiency combining control variate variances and replication costs are consided.
A significance statement is not available in the OpenAlex record.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
Variance Reduction(VR) is an important way to raise simulation estimation efficiency,and Control Variate(CV) estimators are widely used for variance reduction.Based on much works in developing and analyzing CV estimators,some further researches are given.Control variate estimation is studied where the control mean itself is unbiased estimated,and take both prior and auxiliary simulation replications into consideration.Analyzing control variate estimator is focus on where the control variance is known,and given a compound estimator depending on simulation replications and covariances.Then the efficiency combining control variate variances and replication costs are consided.
Key concepts: Control variates, Estimator, Variance reduction, Random variate, Variance (accounting), Replication (statistics), Statistics, Reduction (mathematics)