Bayesians Should Not Resample a Prior Sample to Learn about the Posterior |
| |
Authors: | Sheldon M. Ross |
| |
Affiliation: | Department of Industrial Engineering and Operations Research , University of California , Berkeley , CA , 94720 , USA |
| |
Abstract: | It is argued in a Bayesian context in which a set of parameter values is simulated from a prior distribution, that the sampling importance resampling algorithm should not be used to resample these values so as to obtain an approximate sample from the posterior. Rather, the whole set of prior values, along with their appropriate weights, can be more gainfully employed. |
| |
Keywords: | Prior Sampling importance resampling Simulation |
|
|