Is Sequential Monte Carlo different from Markov Chain Monte Carlo?
@petergreen5900
2 жыл бұрын
Good question. The short answer is: yes, though they are both trying to do the same job i.e. generate samples from an arbitrary probability distribution, but they go about it differently. MCMC involves growing a Markov chain whose stationary distribution is equal to the target, SMC is fundamentally based on Importance Sampling. Does that help?
@iidtxbc
2 жыл бұрын
@@petergreen5900 So, SMC is not based on Monte Carlo Markov Chain? and SMC sample independently based on importance sampling but does sampling sequentially?
@petergreen5900
2 жыл бұрын
@@iidtxbc yep 🙂
@iidtxbc
2 жыл бұрын
@@petergreen5900 Thank you very much for your work and kindness and hope your work helps more people out there.
Пікірлер: 9