arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 1006.3656cond-mat.stat-mech
收藏

经典统计力学反问题的信息论解法

Information-theory-based solution of the inverse problem in classical statistical mechanics

Marco D'Alessandro, Francesco Cilloco

摘要:

我们提出了一个程序的径向对分布函数的知识的相互作用势的确定。该方法基于信息论的最大熵原理,采用反蒙特卡罗模拟方案,以相互作用势作为跃迁概率的渐近表达式。高密度单原子流体得到的结果是非常令人满意的,并提供了一个准确的提取的潜力,尽管适度的计算工作。

英文摘要:

We present a procedure for the determination of the interaction potential from the knowledge of the radial pair distribution function. The method, realized inside an inverse Monte Carlo simulation scheme, is based on the application of the Maximum Entropy Principle of information theory and the interaction potential emerges as the asymptotic expression of the transition probability. Results obtained for high density monoatomic fluids are very satisfactory and provide an accurate extraction of the potential, despite a modest computational effort.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑