摘要: The workshop featured eleven talks concerning various issues and applications of computational Bayesian analysis. All presentations were followed by informal, often lively, discussions. In the spirit of a workshop atmosphere, a tw... 展开 The workshop featured eleven talks concerning various issues and applications of computational Bayesian analysis. All presentations were followed by informal, often lively, discussions. In the spirit of a workshop atmosphere, a two hour 'working lunch' was devoted to an open forum on Bayesian computing. The issues raised focused on which computing environments are desirable for Bayesian statistical analyses, and what are the potentials for a Bayesian software package. In the beginning, there was division of opinions about suitable environment for interactive Bayesian analyses and the question of how to start the process. However, an overwhelming majority of participants agreed with the idea that is too early to push for a new statistical package. Instead we should all push for new software to be compatible with a package like S (an AT&T trademark) so that one could use its data handling and graphics capabilities. Furthermore, we should start a Bayesian Bulletin Board which will give news about Bayesian programs as they are developed and a Bayesian Software Data Base which could be accessed via networks popular among academia. This would be done either at Carnegie-Mellon University's Statistics VAX or at The Ohio State University's Math-Stat PYRAMID. Keywords: Artificial intelligence; Bayes inference; Regression and econometrics model. 收起