Arbeitspapier
Computational Efficiency in Bayesian Model and Variable Selection
Large scale Bayesian model averaging and variable selection exercises present, despite the great increase in desktop computing power, considerable computational challenges. Due to the large scale it is impossible to evaluate all possible models and estimates of posterior probabilities are instead obtained from stochastic (MCMC) schemes designed to converge on the posterior distribution over the model space. While this frees us from the requirement of evaluating all possible models the computational effort is still substantial and efficient implementation is vital. Efficient implementation is concerned with two issues: the efficiency of the MCMC algorithm itself and efficient computation of the quantities needed to obtain a draw from the MCMC algorithm. We evaluate several different MCMC algorithms and find that relatively simple algorithms with local moves perform competitively except possibly when the data is highly collinear. For the second aspect, efficient computation within the sampler, we focus on the important case of linear models where the computations essentially reduce to least squares calculations. Least squares solvers that update a previous model estimate are appealing when the MCMC algorithm makes local moves and we find that the Cholesky update is both fast and accurate.
- Sprache
-
Englisch
- Erschienen in
-
Series: Working Paper ; No. 4/2007
- Klassifikation
-
Wirtschaft
Bayesian Analysis: General
Statistical Simulation Methods: General
Model Evaluation, Validation, and Selection
Computational Techniques; Simulation Modeling
- Thema
-
Bayesian Model Averaging
Sweep operator
Cholesky decomposition
QR decomposition
Swendsen-Wang algorithm
- Ereignis
-
Geistige Schöpfung
- (wer)
-
Eklund, Jana
Karlsson, Sune
- Ereignis
-
Veröffentlichung
- (wer)
-
Örebro University School of Business
- (wo)
-
Örebro
- (wann)
-
2007
- Handle
- Letzte Aktualisierung
-
10.03.2025, 11:41 MEZ
Datenpartner
ZBW - Deutsche Zentralbibliothek für Wirtschaftswissenschaften - Leibniz-Informationszentrum Wirtschaft. Bei Fragen zum Objekt wenden Sie sich bitte an den Datenpartner.
Objekttyp
- Arbeitspapier
Beteiligte
- Eklund, Jana
- Karlsson, Sune
- Örebro University School of Business
Entstanden
- 2007
Ähnliche Objekte (12)
Assessing Brain Activity through Spatial Bayesian Variable Selection
Bayesian computational methods
Quasi-Bayesian model selection
Heterogeneous variable selection in nonlinear panel data models: A semiparametric Bayesian approach
Essays on Bayesian econometrics and computational economics
Bayesian Model Selection for hydro-morphodynamic models
Project selection, income smoothing, and Bayesian learning
Project selection, income smoothing, and Bayesian learning
Bayesian feature selection to estimate customer survival
Computational complexity and parallelization in Bayesian econometric analysis
Efficiency, Adverse Selection, and Production
Approximation of bayesian efficiency in experimental choice designs
Assessing Brain Activity through Spatial Bayesian Variable Selection
Bayesian computational methods
Quasi-Bayesian model selection
Heterogeneous variable selection in nonlinear panel data models: A semiparametric Bayesian approach
Essays on Bayesian econometrics and computational economics
Bayesian Model Selection for hydro-morphodynamic models
Project selection, income smoothing, and Bayesian learning
Project selection, income smoothing, and Bayesian learning
Bayesian feature selection to estimate customer survival
Computational complexity and parallelization in Bayesian econometric analysis
Efficiency, Adverse Selection, and Production
Approximation of bayesian efficiency in experimental choice designs
Assessing Brain Activity through Spatial Bayesian Variable Selection
Bayesian computational methods
Quasi-Bayesian model selection
Heterogeneous variable selection in nonlinear panel data models: A semiparametric Bayesian approach
Essays on Bayesian econometrics and computational economics
Bayesian Model Selection for hydro-morphodynamic models
Project selection, income smoothing, and Bayesian learning
Project selection, income smoothing, and Bayesian learning
Bayesian feature selection to estimate customer survival
Computational complexity and parallelization in Bayesian econometric analysis
Efficiency, Adverse Selection, and Production