| 摘要: |
| 首先研究并证明了Tsallis最大熵和约束条件下的Tsallis最大熵原理；其次，针对最大熵方法的逆问题，讨论了贝叶斯参数估计理论中利用Tsallis最大熵原理确定参数的先验概率的逆问题; 对于一些具体的概率分布，根据Tsallis最大熵原理，利用变分的方法，求解出使Tsallis熵达到最大值的约束条件.该类逆问题的解一般不是惟一的，其他分布情况也可按此方法得出。 |
| 关键词: Tsallis最大熵原理 逆问题 约束条件 |
| DOI： |
| 分类号: |
| 基金项目: |
|
| Tsallis Maximum Entropy Principle and Its Inverse Problem |
|
WU Xian-dong，WANG Jia-mei，LI Wan-qing
|
| Abstract: |
| First, we have studied and proved the Tsallis maximum entropy principle and the Tsallis maximum entropy principle under the constraint conditions. Secondly, in view of the inverse problem of the maximum entropy method, the inverse problem of using the Tsallis maximum entropy principle to determine the prior probabilities in the Bayesian parameter estimation theory is discussed. For some specific probability distributions, the Tsallis maximum entropy principle and variation principle are used to solve the constraint conditions that maximize the Tsallis entropy. The solution of this inverse problem is generally not unique, and other distributions can also be obtained by this method. |
| Key words: Tsallis maximum entropy principle inverse problem constraint condition |
