Stata 18 offers several new features for Bayesian analysis. With the new bma suite, you do not need to choose just one model for your analysis—you can borrow information from many models. bma performs Bayesian model averaging to account for model uncertainty in your analysis. Use bmaregress to find out which predictors are important. Perform model choice, inference, and prediction. Use many postestimation commands to explore influential models, model complexity, model fit and predictive performance, sensitivity analysis to the assumptions about importance of models and predictors, and more.
In StataNow™, the new bayes: qreg command fits Bayesian quantile regressions, and the new likelihood option asymlaplaceq() of bayesmh allows you to fit even more Bayesian quantile models, including simultaneous and multilevel quantile regressions. Also in StataNow, the new bayesselect command performs Bayesian variable selection for linear models.
Read more about these features and see them in action:
• Bayesian model averaging
• Bayesian variable selection for linear regression
StataNow• Bayesian quantile regression
StataNow• Bayesian asymmetric Laplace model
StataNow• And more
;)
View all the new features in Stata 18.
