AI RESEARCH
Aligning Validation with Deployment: Target-Weighted Cross-Validation for Spatial Prediction
arXiv CS.LG
•
ArXi:2603.29981v1 Announce Type: new Cross-validation (CV) is commonly used to estimate predictive risk when independent test data are unavailable. Its validity depends on the assumption that validation tasks are sampled from the same distribution as prediction tasks encountered during deployment. In spatial prediction and other settings with structured data, this assumption is frequently violated, leading to biased estimates of deployment risk.