Dear All,
I’m working on a research problem which is been handled by using simple linear regression previously, where predictor is time varying. I want a comparison of this existing model with Bayesian linear regression and compare the parameters in both the conditions. Theoretically I have all the statistical concept but haven’t done before.
Can anyone suggest me If there is any similar exercises? Or Can you suggest If there is any example or any working exercise solution with R codes?
Best Regards,
Abhay
Please go through these links, you might find them helpful.
http://www.rpubs.com/mfondoum/bayesian_linear_regression
https://cran.r-project.org/web/packages/BLR/BLR.pdf
It would depend on how comfortable you are with coding the model up.
rstanarm (https://cran.r-project.org/web/packages/rstanarm/index.html)
is an excellent package with uses Stan as its engine. If you are using non-informative priors then the syntax is also very similar to glm commands in R.
Its fast and have a lot of tutorials online.
Look at MCMCpack, BOA and, bayesm packages all implementable in R! Hope this helps.
Thank you Sir John Ng'ombe , Mohamed Shehata and Sreenath Arekunnath Madathil
The simplest start is indeed the above mentioned rstanarm package. You simply prefix your command with "stan_", so instead of "lm(...)" you write "stan_lm(...)". If you're not familiar with the concept of priors, rstanarm will automatically generate weak informative priors, so you have at least some regularization for your parameters.
- Badges
- Science topic