Designs for crossvalidating approximation models

DOI10.1093/BIOMET/AST034MaRDI QIDQ2870263zbMATH OpenOpenAlexFDO

Authors Qiong Zhang, Peter Z. G. Qian

Publication date 17 January 2014

Published in Biometrika (Search for Journal in Brave)

Full work available at URL https://semanticscholar.org/paper/06f714e89571cbb3e3d9cf3a873178aa87a1e8c6

zbMATH Keywords

computer experimentsdesign of experimentsblack-box functionssurrogate modelsLatin hypercube designsmeta-modelssliced Latin hypercube designs

Mathematics Subject Classification ID

Statistical block designs (62K10) Orthogonal arrays, Latin squares, Room squares (05B15)

Recommendations

• Minimization and estimation of the variance of prediction errors for cross-validation designs
• Model selection via multifold cross validation
• Improvements on Cross-Validation: The .632+ Bootstrap Method
• A comparative study of ordinary cross-validation, v-fold cross-validation and the repeated learning-testing methods
• Multiple predicting \(K\)-fold cross-validation for model selection

Cited in

(6)

• A simple approach to constructing quasi-Sudoku-based sliced space-filling designs
• Self-supervised cross validation using data generation structure
• Minimization and estimation of the variance of prediction errors for cross-validation designs
• Construction of (nearly) orthogonal sliced Latin hypercube designs
• Sliced Rotated Sphere Packing Designs
• Purposeful cross-validation: a novel cross-validation strategy for improved surrogate optimizability

This page was built for publication: Designs for crossvalidating approximation models

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2870263)