Journal of Chemometrics · 1992 · 121 citations · 15 references
ComputingData SciencePredictive AnalyticsStatistical LearningStatistical ModelingFactorial DesignsFeature SelectionModel SelectionPredictive ModelingVariable SelectionPredictive Model AssessmentPredictive Model SelectionPredictive PerformancePredictive Performance EvaluationRegression Model Selection
Abstract A procedure called GOLPE is suggested in order to detect those variables which increase the predictivity of PLS models. The procedure is based on evaluating the predictive power of a number of PLS models built by different combinations of variables selected according to a factorial design strategy. Examples are given of the efficiency of this variable selection procedure, which shows how these predictive PLS models are better than those obtained by all variables and better than the corresponding ordinary regression models.
15