Record number :
Title of article :
Analysis of MCMC algorithms for Bayesian linear regression with Laplace errors
Author/Authors :
Choi، نويسنده , , Hee Min and Hobert، نويسنده , , James P.، نويسنده ,
Issue Information :
دوفصلنامه با شماره پیاپی سال 2013
Pages :
From page :
To page :
Abstract :
Let π denote the intractable posterior density that results when the standard default prior is placed on the parameters in a linear regression model with iid Laplace errors. We analyze the Markov chains underlying two different Markov chain Monte Carlo algorithms for exploring π . In particular, it is shown that the Markov operators associated with the data augmentation (DA) algorithm and a sandwich variant are both trace-class. Consequently, both Markov chains are geometrically ergodic. It is also established that for each i ∈ { 1 , 2 , 3 , … } , the i th largest eigenvalue of the sandwich operator is less than or equal to the corresponding eigenvalue of the DA operator. It follows that the sandwich algorithm converges at least as fast as the DA algorithm.
Keywords :
Data augmentation algorithm , eigenvalues , Geometric convergence rate , Markov chain , Markov operator , Monte Carlo , Asymmetric Laplace distribution , Trace-class operator , Sandwich algorithm
Journal title :
Journal of Multivariate Analysis
Journal title :
Journal of Multivariate Analysis
Serial Year :
Link To Document :