| |
| | Maximum Expected Utility by MCMC Jacquier (via CobWeb/3.1 planetlab2.ece.ucdavis.edu) (Site not responding. Last check: 2007-10-14) |
 | | MEU problems require both integration, to compute the expected utility, and optimization, to find the optimal decision. |
 | | In most cases of interest, the expected utility does not have a analytical solution, even for a given value of the decision. |
 | | The resulting MCMC algorithm is very efficient, nearly reducing to a pure Gibbs sampler, due to the properties of utility kernels. |
| neumann.hec.ca.cob-web.org:8888 /pages/eric.jacquier/papers/abs.mcmceu.200605.html (159 words) |
|