.png){kind=icon}
.png){kind=icon}
Return
Cross-Laboratory Generalization Failure in Perovskite Solar Cell Machine Learning: A Diagnostic Protocol and Evaluation Threshold
.png){kind=icon}
OA
AI
M
J
H
E
H
D
J
S
J
DOI：10.1016/j.egyai.2026.100856.png)
.png){kind=icon}
.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
• Within-lab R² in perovskite machine learning can conceal cross-lab failure. • Within-lab R² of 0.44 collapses to leave-one-lab-out R² of −3.04 in 819 cells. • The collapse is reproducible across five model families and three laboratories. • A diagnostic protocol with explicit cross-lab threshold (LOLO R² > 0) is provided. • Lab-aware evaluation must become the default, not an optional addition.
Keywords:
Perovskite solar cells
Machine learning
Cross-laboratory generalization
Leave-one-laboratory-out cross-validation
SHAP analysis
AI Summary
Key information extracted from the uploaded paper, including a brief overview, abstract, background, key highlights, visual analysis, and future outlook.
Journal
IF:
9.6
Papers:
835
Citations:
3.1K
Organization
View more organizations