Researchers have developed a new framework for multi-source evidence fusion within the Dempster-Shafer theory, addressing key limitations in existing methods. The proposed system introduces a novel chaos-conflict measurement to jointly assess inter-evidence conflict and intra-evidence uncertainty. Additionally, it incorporates a historical experience-driven weighting scheme that leverages spectral clustering and regret theory to determine the long-term reliability of evidence sources across different decision contexts. Experiments on 16 benchmark datasets show this framework achieves an average F1 score of 85.78 and a mean AUC of 93.30, outperforming several baseline methods. AI
IMPACT This research offers a more robust approach to evidence fusion, potentially improving decision-making in complex systems that rely on integrating data from multiple sources.
RANK_REASON Academic paper detailing a new methodology in AI. [lever_c_demoted from research: ic=1 ai=1.0]
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →
