Parallel Monte Carlo for entropy-robust estimation

MaRDI QIDQ2788878zbMATH OpenFDO

Authors Yuri S. Popkov, Alexey Yu. Popkov, B. S. Darkhovskij

Publication date 22 February 2016

Published in Matematicheskoe Modelirovanie (Search for Journal in Brave)

Full work available at URL http://mathnet.ru/eng/mm3606

zbMATH Keywords

distributionsMonte Carlo methodsparallel algorithmsrandomized algorithms

Mathematics Subject Classification ID

Monte Carlo methods (65C05) Parallel numerical computation (65Y05) Estimation and detection in stochastic control theory (93E10)

Review text

The problem of a non-parametric estimation of the probability density distribution for stochastic dynamical models characteristics with structural non-linearities of a small volume of samples is considered. It is found that optimal probability density distributions are related with an exponential class whose parameters are Lagrange multipliers. A system of equations containing integral components is developed to find the multipliers. The system is solved by parallel algorithms on the base of the Monte Carlo method.

Recommendations

• Estimation and optimization of MLE maximum entropy probability density
• Efficient numerical approximation of maximum entropy estimates
• Estimation of probability density on the basis of the method of stochastic regularization
• Estimating the characteristics of randomized dynamic data models (the entropy-robust approach)
• Asymptotic efficiency of maximum entropy estimates

Cited in

(2)

• Mathematical methods of randomized machine learning
• Estimating the characteristics of randomized dynamic data models (the entropy-robust approach)

This page was built for publication: Parallel Monte Carlo for entropy-robust estimation

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2788878)