A theorem for calculation of the appropriate sample size in an estimation. (November 2017)

• Record Type: Journal Article
• Authors: Zhang, Xue-feng; Yang, Feng-bao; Wang, Xu-zhu
• Is Part Of: Chaos, solitons and fractals. Volume 104(2017)
• Page Start: 291
• Page End: 297
• Publication Date: 2017-11
• Subjects: Fuzzy statistics and data analysis -- Possibility theory -- Simple size for estimation -- Fuzzy decision making -- Credibility distribution; Chaotic behavior in systems -- Periodicals; Solitons -- Periodicals; Fractals -- Periodicals; Chaotic behavior in systems; Fractals; Solitons; Periodicals; 003.7
• DOI: 10.1016/j.chaos.2017.08.015
• Languages: English
• ISSNs: 0960-0779

Abstract

Abstract: In 1982, Dubois and Prade investigated the relationship between belief function, plausibility function and basic probability assignment when the involved universe is finite. In this paper, the similar results on their relationships are obtained with a continuous universe. As an important facility to connect possibility distribution in continuous universes and discrete probability values, basic probability histogram is defined by means of measurement amplitude, which is a notion with both probability and possibility features. A theorem about how to calculate a suitable sample size for estimation is proposed based on the researches on basic probability histograms. Through the theorem, we can directly calculate the appropriate samples size for any population distribution. Even with small samples, a reasonable estimation can be obtained with a non-normal distribution.