Premium
Statistical Tests to Validate Predictive Models
Author(s) -
Hamada M. S.,
Abes J. I.
Publication year - 2015
Publication title -
quality and reliability engineering international
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.913
H-Index - 62
eISSN - 1099-1638
pISSN - 0748-8017
DOI - 10.1002/qre.1715
Subject(s) - replicate , predictive power , model validation , computer science , statistical hypothesis testing , validation test , predictive modelling , test (biology) , statistics , reliability engineering , data mining , engineering , machine learning , mathematics , test validity , psychometrics , philosophy , data science , epistemology , paleontology , biology
This article presents validation tests to check a predictive model over time. These validation tests are like individuals,X ̄ and S control charts. The individuals validation test is used when there are no replicates at the checked time points. TheX ̄ and S validation tests can be used when there are replicates. Their power is evaluated under various scenarios via a simulation study. Based on these results, theX ̄ and S validation tests are recommended when there are replicate measurements. Copyright © 2014 John Wiley & Sons, Ltd.