I want to sample from a distribution $p(X)$ with $X \in R^n$. However, I can only evaluate the likelihoods of $Z = AX$ and $Z = BX$ with $A,B \in R^{m \times n}$ and $m = n-1$.

Now my idea is to run two Markov Chain Monte Carlo samplers in parallel while minimizing the Kullback-Leibler divergence of the sample-approximated distributions of both chains.

I was thinking of some kind of autoregressive proposal distribution that becomes tighter when the approximated distribution of a chain is wider than that of the other chain (and vice versa).

Is this feasible? Does anyone know an existing algorithm that does something similar? I would be very happy about any expert opinion on that.


Your Answer

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

Browse other questions tagged or ask your own question.