Article
Engineering, Aerospace
Xiaojing Fan, Deqiang Han, Jean Dezert, Yi Yang
Summary: In information fusion, it is crucial to construct the transformation between different frameworks when fusing uncertain information modeled with different theoretical frameworks.
CHINESE JOURNAL OF AERONAUTICS
(2023)
Article
Computer Science, Information Systems
Chengxi Yang, Fuyuan Xiao
Summary: Negation is a crucial operation in evidence theory, but in complex evidence theory, which is based on complex number field, negation is still an open problem. Therefore, a new negation method called CBBA exponential negation is proposed and discussed in this paper. This method transforms a CBBA to another one with increased entropy. Various properties of this negation method are rigorously proved, and its impact on negation convergence is also studied. Additionally, a new entropy for CBBA and some numerical examples are presented, and an application of CBBA exponential negation is demonstrated.
INFORMATION SCIENCES
(2023)
Article
Mathematics, Interdisciplinary Applications
Yige Xue, Yong Deng
Summary: Dempster-Shafer evidence theory is an extension of classical probability theory used in the evidential environment. The decomposable entropy in this theory has high theoretical and application value, and the article proposes a new decomposable Deng entropy that effectively decomposes Deng entropy. Experimental results demonstrate the efficiency of the proposed model in decomposing Deng entropy.
CHAOS SOLITONS & FRACTALS
(2022)
Article
Physics, Multidisciplinary
Mohammad Reza Kazemi, Saeid Tahmasebi, Francesco Buono, Maria Longobardi
Summary: The paper introduces the fractional versions of Deng entropy and extropy in DST, comparing them to other measures and studying their maximum. Through an analysis of a classification problem, the importance of these new measures is highlighted.
Article
Computer Science, Information Systems
Shengjia Zhang, Fuyuan Xiao
Summary: Complex evidence theory, as a generation model of the Dempster-Shafer evidence theory, can express and reason uncertainty. The complex basic belief assignment (CBBA) generation method is a key issue in this theory, and modeling uncertainty information remains an open issue. In this paper, a CBBA generation method utilizing triangular fuzzy numbers is proposed, and a decision-making algorithm based on this method is developed. The effectiveness of the algorithm is verified through its application in classification. Overall, the proposed method offers a promising approach for uncertainty modeling and reasoning in both the real number domain and the complex number domain in decision making theory.
INFORMATION SCIENCES
(2023)
Article
Computer Science, Artificial Intelligence
Xiaoyu Han, Xiubin Zhu, Witold Pedrycz, Zhiwu Li
Summary: This study designs a three-way classification mechanism by combining fuzzy decision trees and expressing uncertainty. A fuzzy decision tree is constructed through generalization and the three-way decision model is widely used. An efficient way to flag uncertain data is proposed, which is not possible with commonly used fuzzy decision trees. The developed mechanism consists of two stages: building a fuzzy decision tree and determining the uncertainty level to reject instances. The rejection quality is quantified in terms of accuracy and coefficient, and the mechanism performs better than other three-way decision models.
APPLIED SOFT COMPUTING
(2023)
Article
Physics, Multidisciplinary
Kangkai Gao, Yong Wang, Liyao Ma
Summary: In this paper, a new decision tree method based on belief entropy is proposed, and then extended to random forest. This method can handle continuous attribute values directly without discretization preprocessing, and exhibits good classification accuracy on UCI dataset, especially in situations with high uncertainty.
Article
Computer Science, Artificial Intelligence
Sebastian Porebski
Summary: This paper proposes a novel technique called EasIeR for linguistic rule extraction, which focuses on improving the explainability and reliability of decision support systems.
EXPERT SYSTEMS WITH APPLICATIONS
(2022)
Article
Computer Science, Artificial Intelligence
Xiaowei Gu, Plamen P. Angelov, Qiang Shen
Summary: In this article, a novel zero-order EFS model with a unique belief structure is proposed for data stream classification. The model can handle interclass overlaps and better capture the underlying structure of data streams through prototypes. Experimental results demonstrate the superior performance of the model on various classification problems.
IEEE TRANSACTIONS ON FUZZY SYSTEMS
(2022)
Article
Computer Science, Artificial Intelligence
Lipeng Pan, Yong Deng
Summary: The main contribution of this paper is to propose a complex-valued Deng entropy for measuring uncertainty in the complex-valued evidence theory. The complex-valued Deng entropy is a generalization method that effectively measures uncertainty in the complex-valued framework, considering phase angle information, inconsistency, and non-specificity. Numerical examples demonstrate the compatibility and effectiveness of the complex-valued Deng entropy. Results from classification experiments reveal that the core parameter of the complex-valued Deng entropy produces more accurate results in certain datasets compared to the core parameter of the Deng entropy.
APPLIED INTELLIGENCE
(2023)
Article
Mathematics, Interdisciplinary Applications
Zihan Yu, Yong Deng
Summary: This paper proposes a method to derive power law distribution with maximum Deng entropy and illustrates the properties of the distribution through numerical examples.
CHAOS SOLITONS & FRACTALS
(2022)
Article
Computer Science, Artificial Intelligence
Haojian Huang, Zhe Liu, Xue Han, Xiangli Yang, Lusi Liu
Summary: DST theory is popular due to its advantages in managing uncertain information, but it may produce counterintuitive results when faced with conflicting evidence. To address this flaw, a new belief logarithmic similarity measure (εBLSM) based on DST is proposed. An enhanced belief logarithmic similarity measure (εBLSM) is also presented to consider subset discrepancies. Moreover, a new multi-source data fusion method based on εBLSM is devised, which demonstrates rationality and effectiveness in fault diagnosis and target recognition cases.
JOURNAL OF INTELLIGENT & FUZZY SYSTEMS
(2023)
Article
Mathematics, Interdisciplinary Applications
Omid Kharazmi, Javier E. Contreras-Reyes
Summary: The purpose of this work is to introduce Deng-Fisher information (DFI), Deng-Fisher information distance (DFID), and Jensen-Deng-Fisher (JDF) information distance measures based on the basic probability assignment concept. Results associated with these measures are presented, and DFI and DFID measures for escort of basic probability assignment functions are examined. The Game of Life cellular automaton by Conway is used for illustration, and numerical results based on the proposed information measures are presented, indicating that JDF information distance measures the dynamics of living cell population along time.
CHAOS SOLITONS & FRACTALS
(2023)
Article
Automation & Control Systems
Jixiang Deng, Yong Deng
Summary: This paper proposes dynamic belief entropy (DBE) which improves the measurement of uncertainty by introducing a dynamic parameter. A dynamic data fusion method based on DBE is designed, enhancing the classification performance. Experimental results demonstrate the effectiveness of the proposed method.
ENGINEERING APPLICATIONS OF ARTIFICIAL INTELLIGENCE
(2023)
Article
Mathematics, Interdisciplinary Applications
Siran Li, Fuyuan Xiao
Summary: We define the distribution of uncertainty using Deng entropy and prove its limiting form to be the normal distribution through the characteristic function. We discuss the distribution, analyze the error of its normal approximation, and provide conditions for required error. Furthermore, we reveal the relationship between the distribution and the binomial distribution, highlighting the advantages of using it as a prior distribution in evidence theory.
CHAOS SOLITONS & FRACTALS
(2023)