Konferansepresentasjoner

Modified jack-knifing in multivariate regression for variable selection.

Årstall	2000
Abstract	Introduction One of the criticisms of PLS regression has been the lack of significance tests of the model parameters. Up to now "rules of thumb" exist, mainly based on previous experience when selecting significant wavelengths. We present a novel method based on jack-knifing to estimate variances of model parameters in PCR and PLS regression for cross-validation.The purpose of this work is to develop a model that predicts ethanol concentrations in mixtures of methanol, ethanol and propanol from near infrared (NIR) absorbance data with different techniques.
Referanse	Westad, F., Byström, M., Martens, H. 2000. Modified jack-knifing in multivariate regression for variable selection. Proc. 9th Intl. Conference NIR spectroscopy, 13-19 June 1999, Near Infrared Spectroscopy,,,, Verona, Italy, pp 247-251. ISBN 0 9528666 1 7
Utgiver	Near Infrared Spectroscopy,,,,