Convex evidence theory is the only way to handle ordered and fuzzy evidence fusion. However, conventional convex evidence theory has some drawbacks that make the fusion results unreasonable in some cases, and not efficient in the scenario of massive data. To overcome above issues, in this article we proposed a novel convex evidence theory based on Gaussian functions. We modified Gaussian functions and use them to combine the mass function of ordered propositions. We design the formula of the parameters of Gaussian functions, and propose a more accurate method to find the most likely true proposition. We also prove the effectiveness of the proposed method. Theoretical analysis and experimental results demonstrate that the proposed method has lower time complexity and higher accuracy than state-of-the-art methods.
In JIFS'17, the Journal of Intelligent & Fuzzy Systems, Volume Preprint, pages 1-8, 2017.
This work was partially supported by the AFOSR FA9550-14-1-0261, NSF-Frontiers
Cyber-Physical Heart Award, FWF-NFN RiSE Award, FWF-DC LMCS Award, FFG Harmonia
Award, FFG Em2Apps Award, and the TUW CPPS-DK Award.