Likelihood weighting is a version of “importance sampling” where we fix the evidence variable to (we only sample relevant events). We have to weight them with the likelihood that they fit the evidence and use the probabilities we know to do so.
Likelihood weighting is a version of “importance sampling” where we fix the evidence variable to (we only sample relevant events). We have to weight them with the likelihood that they fit the evidence and use the probabilities we know to do so.