Agentic data science with the Data Agent Kit
cloud.google.com·6h ago
AI Research
arXiv CS.AI·Aug 15·5 min read
Robust Dempster-Shafer Evidence Fusion with Chaos-Conflict Measurement and Historical-Experience Weighting
TL;DR
A new framework for Dempster-Shafer evidence fusion integrates chaos-conflict measurement and historical-experience weighting, achieving superior classification performance. It outperforms existing methods with an average F1 score of 85.78 and a mean AUC of 93.30 across multiple datasets.
✦ Why It Matters
Implement this framework to enhance decision-making accuracy in multi-source data environments today.
Key Takeaways
- Utilize chaos-conflict measurement for better evidence assessment.
- Apply historical experience weighting for context-specific reliability.
- Leverage hybrid combination rules for adaptive decision-making.
- Test the framework on diverse datasets to validate performance.
How It Works
The framework employs chaos-conflict measurement to jointly assess conflicts between different evidence sources and the uncertainty within each source. This is complemented by a historical experience weighting scheme that categorizes decision contexts using spectral clustering, allowing for the calculation of reliability profiles based on past fusion results.
The combination of these methods leads to a hybrid rule that dynamically adjusts the balance between preserving uncertainty and achieving consensus, ultimately enhancing classification robustness.