Bayesian model averaging in R

Amini, Shahram; Parmeter, Christopher F.

doi:10.3233/jem-2011-0350

Cited by 46 publications

(26 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The BMS package implements Bayesian model averaging for R. It excels in offering a range of widely used prior structures coupled with efficient MCMC algorithms to sort through the model space. Amini and Parmeter (2011) and Amini and Parmeter (2012) carry out a comparison of R software packages that implement Bayesian model averaging, in particular the packages BAS (Clyde 2012) and BMA (Raftery, Hoeting, Volinsky, Painter, and Yeung 2014). Amini and Parmeter (2012) conclude that package BMS is the only one among its competitors that is able to reproduce empirical results in Fernández et al (2001b); Doppelhofer and Weeks (2009) and the working paper version of Masanjala and Papageorgiou (2008).…”

Section: Discussionmentioning

confidence: 99%

Bayesian Model Averaging Employing Fixed and Flexible Priors: TheBMSPackage forR

Zeugner¹,

Feldkircher²

2015

J. Stat. Soft.

174

164

View full text Add to dashboard Cite

This article describes the BMS (Bayesian model sampling) package for R that implements Bayesian model averaging for linear regression models. The package excels in allowing for a variety of prior structures, among them the "binomial-beta" prior on the model space and the so-called "hyper-g" specifications for Zellner's g prior. Furthermore, the BMS package allows the user to specify her own model priors and offers a possibility of subjective inference by setting "prior inclusion probabilities" according to the researcher's beliefs. Furthermore, graphical analysis of results is provided by numerous built-in plot functions of posterior densities, predictive densities and graphical illustrations to compare results under different prior settings. Finally, the package provides full enumeration of the model space for small scale problems as well as two efficient MCMC (Markov chain Monte Carlo) samplers that sort through the model space when the number of potential covariates is large.

show abstract

Section: Discussionmentioning

confidence: 99%

Bayesian Model Averaging Employing Fixed and Flexible Priors: TheBMSPackage forR

Zeugner¹,

Feldkircher²

2015

J. Stat. Soft.

174

164

View full text Add to dashboard Cite

show abstract

“…The growth in options for researchers wishing to conduct statistical analysis using BMA is rapid. When Amini and colleagues (2011) [ 44 ] conducted a review of BMA libraries in R there were three (3) official libraries available on CRAN: BAS [ 53 ], BMA [ 51 ] and BMS [ 56 ]. Current libraries available through CRAN, which can be viewed at https://cran.r-project.org/view=Bayesian , include updated versions of these packages, and in addition, libraries such as mlogitBMA [ 57 ], BayesVarSel [ 58 ] and BoomSpikeSlab [ 59 ].…”

Section: Discussionmentioning

confidence: 99%

“…A more detailed discussion of Bayesian model selection for social scientists is given by [ 21 , 42 ]. [ 43 , 44 ] provide practical tutorials on implementing the BMA using the “BMA” R package.…”

Section: Determining the Significance Of The Coefficients In Generalimentioning

confidence: 99%

Non-significant p-values? Strategies to understand and better determine the importance of effects and interactions in logistic regression

Vakhitova

Alston-Knox

2018

PLoS ONE

View full text Add to dashboard Cite

In the context of generalized linear models (GLMs), interactions are automatically induced on the natural scale of the data. The conventional approach to measuring effects in GLMs based on significance testing (e.g. the Wald test or using deviance to assess model fit) is not always appropriate. The objective of this paper is to demonstrate the limitations of these conventional approaches and to explore alternative strategies for determining the importance of effects. The paper compares four approaches to determining the importance of effects in the GLM using 1) the Wald statistic, 2) change in deviance (model fitting criteria), 3) Bayesian GLM using vaguely informative priors and 4) Bayesian Model Averaging analysis. The main points in this paper are illustrated using an example study, which examines the risk factors for cyber abuse victimization, and are further examined using a simulation study. Analysis of our example dataset shows that, in terms of a logistic GLM, the conventional methods using the Wald test and the change in deviance can produce results that are difficult to interpret; Bayesian analysis of GLM is a suitable alternative, which is enhanced with prior knowledge about the direction of the effects; and Bayesian Model Averaging (BMA) is especially suited for new areas of research, particularly in the absence of theory. We recommend that social scientists consider including BMA in their standard toolbox for analysis of GLMs.

show abstract

“…For example, Baumeister et al (2017) pointed out that different models outperformed others over different forecast horizons, and they claimed that using simple average of predictions from all models yielded the best forecasts. The simple average method can be treated as the simplest format of ensemble learning, whose popular methods applied include boosting (Freund & Schapire, 1997;Mason, Baxter, Bartlett, & Frean, 2000), bagging (Breiman, 1996), stacking (Smyth & Wolpert, 1999;Clarke, 2003), and Bayesian model average (Hoeting, Madigan, Raftery, & Volinsky, 1999;Amini & Parmeter, 2011). Taking the Bayesian model average (BMA) method for example, while assuming that the frequencies of k proposed models to be selected for the prediction purposes follow a multinomial distribution with parameters p i 's, a priori a conjugate Dirichlet distribution can be assigned to those p i 's, denoted as…”

Section: Time-series Prediction Models For Gasoline Pricesmentioning

confidence: 99%

Time-Series Forecasting Models for Gasoline Prices in China

Xu¹,

Sepehri²,

Ma³

et al. 2018

IJEF

View full text Add to dashboard Cite

Accurate prediction of gasoline price is important for the automobile makers to adjust designs and productions as well as marketing plans of their products. It is also necessary for government agencies to set effective inflation monitoring and environmental protection policies. To predict future levels of the gasoline price, due to difficulties of obtaining accurate estimates of influential external factors, data driven time-series forecasting models thus become more suitable given the convenience and practicability they are providing. In this paper, five popular time-series forecasting models, i.e., ARIMA-GARCH, exponential smoothing, grey system, neural network, and support vector machines models, are applied to predict gasoline prices in China. Comparing the performances of these models, it is noted that for this specific time series, a parsimonious ARIMA model performs the best in predicting the gasoline prices for a short time horizon, while for the medium length and long run the SVR and FNN models outperforms others respectively.  

show abstract

Bayesian model averaging in R

Cited by 46 publications

References 16 publications

Bayesian Model Averaging Employing Fixed and Flexible Priors: TheBMSPackage forR

Bayesian Model Averaging Employing Fixed and Flexible Priors: TheBMSPackage forR

Non-significant p-values? Strategies to understand and better determine the importance of effects and interactions in logistic regression

Time-Series Forecasting Models for Gasoline Prices in China

Contact Info

Product

Resources

About