I'm intending to implement a Bayesian LASSO inside the Gibbs sampler I use to estimate a multivariate time-series model, but I have a doubt about how to draw this step.

The prior is a Double-Exponential (a.k.a. Laplace distribution) and the likelihood is Gaussian. Unlike Gaussian likelihood with a Gaussian prior, this choice does not seem to conjugate and I can't see a way to use inversion method. How do I draw from this combination? Should I use an approximation like Metropolis-Hastings?

Thanks in advance.


I assume that the Laplace prior is on the mean of the Gaussian random variable.

Sadly yes, Gaussian likelihood and Laplace prior do not yield a tractable posterior distribution. You will have to use some MCMC method such as Metropolis-Hastings or slice sampling.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.