THE APPLICATION OF THE MAXIMUM ENTROPY PRINCIPLE IN THE FORECAST FOR PROBABILITY DISTRIBUTION

Aiping FANG,Pengbo TIAN,Yi JIA,Youli YU,Hongrong LI,Xiaoli WANG

DOI: https://doi.org/10.3969/j.issn.1009-7104.2017.06.019

2017-01-01

Abstract:By the theory of information entropy,this paper gives general statements of maxi-mum entropy problem.With maximum entropy principle,it is possible to make global j udg-ments and decisions in the absence of major information.Taken the example of probability dis-tribution of divergent rings in shooting sports,the linear programming (LP)is solved based on Lingo,and the limitation of the high-score rings in this sport has been discussed.After showing the variety of maximum entropy in divergent mean value of rings,the data reveal the uniformity and symmetry of probability under different maximum entropy.From this simple model,we can conclude that entropy can demonstrate the consistency of probability distribu-tions,and large entropy generally corresponds to high uniformity in probability.