A multi-source information fusion method based on a new base belief function
Palash Dutta, Sonom Shome
Dibrugarh University
阅读操作
确认中在文库中上传 PDF 后可生成中文音频讲解。
摘要与影响
Dempster-Shafer (D-S) evidence theory has been widely used in applications such as decision-making and pattern recognition. However, the classical Dempster combination rule (DCR) often produces results that do not accurately reflect the true belief distribution, particularly when the input evidence is highly conflicting. To address this issue, this paper introduces a base belief function designed to adjust the basic probability assignment (BPA) before combination, within a closed-world framework. The proposed approach aims to produce more intuitive and reasonable fusion results, especially in situations involving significant evidence conflict. In real-world scenarios, sources of information are frequently unreliable due to differing levels of access to knowledge and varied expertise. Experts may rely on diverse backgrounds, methodologies, and experiences, inevitably leading to inconsistencies and conflicts during evidence fusion. Moreover, DCR enforces a unity property through compulsive normalization of fusing results, which can exacerbate conflicts and yield counterintuitive outcomes. To effectively manage potential conflicts in multisource information fusion systems, this paper proposes a novel information combination approach grounded in evidence correction, reliability assessment, and conflict redistribution. First, by employing the proposed base belief function, evidence is corrected to prevent null BPA conditions. Then, using a reliability degree derived from a distance function, a weighted average BPA is calculated, addressing conflicts arising from unreliable sources. Additionally, instead of directly applying DCR, a conflict redistribution strategy is implemented based on the conflict factor between the corrected evidence and the weighted BPA. This approach avoids the issues caused by forced normalization. Overall, the conflict management strategy based on the base belief function not only ensures the rationality and validity of fusion results but also improves the reliability and robustness of multi-source fusion systems. The effectiveness and superiority of the proposed method are demonstrated through two numerical experiments involving a high level of conflict, based on both single-subset and multi-subset focal elements.
逐年被引趋势
暂无年度引用数据
关键指标
同类平均 = 1
同领域 · 同年份 · 同类型
Google Scholar 与 OpenAlex 的被引统计范围不同,数值存在差异属正常。
AI 辅助阅读
依据:摘要
可就本文提问;依据不足时会说明。
学术脉络
学科主题
计算机 / AIMulti-Criteria Decision Making
Advanced Decision-Making Techniques · Fuzzy and Soft Set Theory
参考文献 94
此处列出前 3 条