[arXiv]score: 0.31
Physics-Guided ML Infers CO Concentration from SnO-SnO₂ Sensor Transients Across p-n Switching Regimes
May 26, 2026
A physics-guided ML framework using physically interpretable transient descriptors combined with FFT/DWT features and leakage-aware cross-validation achieves strong CO concentration inference from a mixed-phase gas sensor exhibiting temperature-dependent p-n switching.
physics.chem-phcs.LGphysics.app-ph
HOW THIS AFFECTS YOU
●
researcherThe leakage-aware grouped cross-validation design and physics-descriptor approach provide a replicable methodology for ML on sensor data with complex regime-switching behavior.