Success Probability Assessment Based on Information Entropy

  • Xuan Chen National University of Defense Technology, Changsha
  • Hanyan Huang National University of Defense Technology, Changsha
  • Zhengming Wang National University of Defense Technology, Changsha
Keywords: Information entropy, equivalent surrogate test, equivalent source, Bayesian method, Bootstrap method

Abstract

The Bayesian method is superior to the classical statistical method on condition of small sample test. However, its evaluation results are not so good if subjective prior information is intervened. The success probability assessment about the success or failure tests of weapon products focussed in this paper, and a fusing evaluation method based on information entropy is proposed. Firstly, data from equivalent surrogate tests is converted into the prior information of an equivalent source by the information entropy theory. Secondly, the prior distribution of the success probability is identified via the Bootstrap method, and the posterior distribution is provided by the Bayesian method with the information of prototype tests in succession. Lastly, an example is given, and the results show that the proposed method is effective and valuable.

Defence Science Journal, 2010, 60(3), pp.271-275, DOI:http://dx.doi.org/10.14429/dsj.60.353

Author Biographies

Xuan Chen, National University of Defense Technology, Changsha
His research areas include: test data processing, test design, and data fusion.
Hanyan Huang, National University of Defense Technology, Changsha
Her research interests include: test data processing, experimental design, and data fusion.
Zhengming Wang, National University of Defense Technology, Changsha
Dean of the College of Science at National University of Defense Technology. His research interests include: image processing, test data processing, experimental design, and data fusion.
Published
2010-04-21
How to Cite
ChenX., HuangH., & WangZ. (2010). Success Probability Assessment Based on Information Entropy. Defence Science Journal, 60(3), 271-275. https://doi.org/10.14429/dsj.60.353
Section
Review Papers