Likelihood Ratio Gradient Estimation for Steady-State Parameters
نویسندگان
چکیده
منابع مشابه
Likelihood Ratio Gradient Estimation for Steady-State Parameters
Abstract: We consider a discrete-time Markov chain Φ on a general state-space X, whose transition probabilities are parameterized by a real-valued vector θ. Under the assumption that Φ is geometrically ergodic with corresponding stationary distribution π(θ), we are interested in estimating the gradient ∇α(θ) of the steady-state expectation α(θ) = π(θ)f. To this end, we first give sufficient con...
متن کاملLikelihood Ratio Gradient Estimationfor Stochastic
In this paper, we develop mathematical machinery for verifying that a broad class of general state space Markov chains reacts smoothly to certain types of perturbations in the underlying transition structure. Our main result provides conditions under which the stationary probability measure of an ergodic Harris recurrent Markov chain is diierentiable in a certain strong sense. The approach is b...
متن کاملMaximum Likelihood Estimation and Likelihood-ratio Tests
The method of maximum likelihood (ML), introduced by Fisher (1921), is widely used in human and quantitative genetics and we draw upon this approach throughout the book, especially in Chapters 13–16 (mixture distributions) and 26–27 (variance component estimation). Weir (1996) gives a useful introduction with genetic applications, while Kendall and Stuart (1979) and Edwards (1992) provide more ...
متن کاملEstimation of steady - state basic parameters of stars
From a minimum of total energy of celestial bodies, their basic parameters are obtained. The steady-state values of mass, radius, and temperature of stars and white dwarfs, as well as masses of pulsars are calculated. The luminosity and giromagnetic ratio of celestial bodies are estimated. All the obtained values are in a satisfactory agreement with observation data.
متن کاملSteady-state quantile estimation
Steady state simulation is used to study long-run behavior. Usually only the expected value of the steady state probability distribution function is estimated. In many cases quantiles of this distribution are of higher interest. In this paper a new usage of quantile estimators is proposed, which is derived from mean value analysis and is based on multiple independent replications. The advantage...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
ژورنال
عنوان ژورنال: Stochastic Systems
سال: 2019
ISSN: 1946-5238,1946-5238
DOI: 10.1287/stsy.2018.0023