A new complex evidence theory

Article Mathematics, Interdisciplinary Applications

Belief entropy-of-entropy and its application in the cardiac interbeat interval time series analysis

Huizi Cui et al.

Summary: This paper introduces a method based on belief entropy to measure the complexity of physiological signals in biological systems. The method has better accuracy and applicability compared to existing entropy algorithms.

CHAOS SOLITONS & FRACTALS (2022)

添加到收藏夹

Article Automation & Control Systems

Random Permutation Set

Yong Deng

Summary: This paper explores the meaning of the power set in evidence theory and proposes a possible explanation of the power set based on Pascal's triangle and combinatorial number. It introduces a new kind of set called random permutation set (RPS), which consists of permutation event space (PES) and permutation mass function (PMF). The paper also discusses and summarizes the comparisons of probability theory, evidence theory, and RPS, and presents an RPS-based data fusion algorithm, which is applied in threat assessment and proves to effectively handle uncertainty.

INTERNATIONAL JOURNAL OF COMPUTERS COMMUNICATIONS & CONTROL (2022)

添加到收藏夹

Article Physics, Multidisciplinary

A unified formulation of entropy and its application

Narayanaswamy Balakrishnan et al.

Summary: This paper proposes a general formulation of entropy, which includes Shannon, Tsallis, and fractional entropy as special cases. The properties of the fractional Tsallis entropy are studied, and a corresponding entropy in the context of Dempster-Shafer theory of evidence, called the fractional version of Tsallis-Deng entropy, is proposed. Finally, an application to two classification problems is presented.

PHYSICA A-STATISTICAL MECHANICS AND ITS APPLICATIONS (2022)

添加到收藏夹

Article Computer Science, Information Systems

Multi-stage consistency optimization algorithm for decision making with incomplete probabilistic linguistic preference relation

Peng Wang et al.

Summary: An optimization algorithm is developed in this study for preference decision-making with incomplete probabilistic linguistic preference relation (InPLPR). By constructing a two-stage mathematical optimization model based on expected multiplicative consistency, missing information can be estimated and consistency improved, ultimately leading to the ranking of alternatives.

INFORMATION SCIENCES (2021)

添加到收藏夹

Article Computer Science, Artificial Intelligence

A consensus model for hesitant fuzzy linguistic group decision-making in the framework of Dempster-Shafer evidence theory

Peide Liu et al.

Summary: This paper introduces a new consensus building process model for group decision-making with hesitant fuzzy linguistic term sets, taking into account the reliability of experts and their weights. By calculating expert reliability and providing explicit advice, the discourse power of experts is strengthened, helping the group achieve consensus.

KNOWLEDGE-BASED SYSTEMS (2021)

添加到收藏夹

Article Computer Science, Information Systems

Conflicting evidence combination from the perspective of networks

Leihui Xiong et al.

Summary: This paper proposes a new evidence combination method based on the interaction among nodes in complex networks, which combines evidence efficiently using weighted average and Dempster's rule of combination, showing better performance through a numerical example.

INFORMATION SCIENCES (2021)

添加到收藏夹

Article Computer Science, Artificial Intelligence

Novel weighted ensemble classifier for smartphone based indoor localization

Priya Roy et al.

Summary: Indoor localization systems using WiFi signals face challenges due to the significant variation of signal strength with ambient conditions and device configuration. This paper proposes a weighted ensemble classifier based on Dempster-Shafer belief theory to efficiently handle context heterogeneity. Real life experiments show that the technique achieves high localization accuracy at varying granularity levels.

EXPERT SYSTEMS WITH APPLICATIONS (2021)

添加到收藏夹

Article Computer Science, Information Systems

Improving the robustness of recursive consequent parameters learning in evolving neuro-fuzzy systems

Edwin Lughofer

Summary: Evolving (neuro-) fuzzy systems have gained popularity in data stream mining and modeling processes due to their ability to update in real-time and adjust models to process drifts. Alternative variants for consequent parameter updates, such as multi-innovation RFWLS and recursive correntropy, outperform standard RFWLS in practical applications and time-series forecasting, showing lower sensitivity to data noise levels.

INFORMATION SCIENCES (2021)

添加到收藏夹

Article Computer Science, Artificial Intelligence

A distance for belief functions of orderable set

Cuiping Cheng et al.

Summary: This paper introduces a method for measuring conflicts between ordered sets, which effectively quantifies the similarity between sets and provides a more robust and accurate characterization of the agreement of ordered sets.

PATTERN RECOGNITION LETTERS (2021)

添加到收藏夹

Article Computer Science, Artificial Intelligence

CED: A Distance for Complex Mass Functions

Fuyuan Xiao

Summary: Evidence theory is an effective methodology for handling uncertainty and a number of distance measures have been proposed to represent the difference between pieces of evidence. This article introduces a generalized evidential distance measure called Complex Evidential Distance (CED) for complex evidence theory, providing a promising way to measure differences between evidence in a more general framework. The CED is a strict distance metric that satisfies distance axioms and has greater ability to measure differences between pieces of evidence, even extending to complex basic belief assignments.

IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS (2021)

添加到收藏夹

Article Computer Science, Artificial Intelligence

Combination of Transferable Classification With Multisource Domain Adaptation Based on Evidential Reasoning

Zhun-Ga Liu et al.

Summary: In this paper, a decision-level combination method for multisource domain adaptation based on evidential reasoning is proposed to improve classification accuracy by considering the reliabilities/weights of different source domains. Dempster's rule is employed to combine the classification results from different domains, and a neighborhood-based cautious decision-making rule is introduced to reduce error risk. Through these methods, the classification accuracy can be efficiently improved and the partial imprecision of classification can be well characterized.

IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS (2021)

添加到收藏夹

Article Computer Science, Artificial Intelligence

Location-Free Robust Scale Estimates for Fuzzy Data

Sara de la Rosa de Saa et al.

Summary: This article discusses the importance of scale measures/estimates in analyzing fuzzy-valued imprecise data and introduces the concept of median distance deviation about the median (MDD) for fuzzy data sets and its robustness. The study points out that calculating MDD in fuzzy data cases is more complex and cannot be precisely computed, but the estimation method for fuzzy trapezoidal data demonstrates a certain level of robustness.

IEEE TRANSACTIONS ON FUZZY SYSTEMS (2021)

添加到收藏夹

Article Computer Science, Artificial Intelligence

Combination of Classifiers With Different Frames of Discernment Based on Belief Functions

Zhunga Liu et al.

Summary: The article proposes a method for classifier fusion based on belief functions to efficiently combine classifiers under different Frames of Discernment (FoD). The method can effectively improve classification accuracy.

IEEE TRANSACTIONS ON FUZZY SYSTEMS (2021)

添加到收藏夹

Article Computer Science, Artificial Intelligence

Fuzzy Hoeffding Decision Tree for Data Stream Classification

Pietro Ducange et al.

Summary: Data stream mining is gaining popularity due to the need for continuous analysis of streaming data, which requires appropriate techniques as traditional machine learning algorithms struggle with fast data streams. This paper introduces a fuzzy Hoeffding Decision Tree (FHDT) that enhances the traditional HDT to be more robust to noisy and vague data. FHDT outperforms HDT, especially in the presence of concept drift, and is highly interpretable thanks to the linguistic rules that can be extracted from it.

INTERNATIONAL JOURNAL OF COMPUTATIONAL INTELLIGENCE SYSTEMS (2021)

添加到收藏夹

Article Computer Science, Artificial Intelligence