Estimation of State Probabilities Using the Maximum Entropy Principle
暂无分享,去创建一个
A simple method is derived for computing state probabilities of a system when the probabilities of certain aggregate states are known. The method is based on maximizing the system entropy. It is shown that the results obtained by the method satisfy certain assumptions on statistical independence between events. The method is applied to a problem arising in computer performance analysis.
[1] S. Kullback,et al. The Information in Contingency Tables , 1980 .
[2] Yonathan Bard,et al. A model of shared dasd and multipathing , 1980 .
[3] A. Katz. Principles of statistical mechanics : the information theory approach , 1967 .
[4] A. E. Ferdinand. A statistical mechanical approach to systems analysis , 1970 .