Hybrid feature-level and model-level fusion of physics residuals with ML ensembles improves CSTR diagnostic accuracy by 2.9% and yields smaller, well-calibrated conformal prediction sets.
Assumption-based design of hybrid diagnosis systems: Analyzing model-based and data-driven principles,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
A Hybrid Intelligent Framework for Uncertainty-Aware Condition Monitoring of Industrial Systems
Hybrid feature-level and model-level fusion of physics residuals with ML ensembles improves CSTR diagnostic accuracy by 2.9% and yields smaller, well-calibrated conformal prediction sets.