Cage of covariance in calibration modeling

Cage of covariance in calibration modeling: Regressing multiple and strongly correlated response variables onto a low rank subspace of explanatory variables

Publikation: Bidrag til tidsskrift › Tidsskriftartikel › Forskning › fagfællebedømt

Standard

Cage of covariance in calibration modeling : Regressing multiple and strongly correlated response variables onto a low rank subspace of explanatory variables. / Eskildsen, Carl Emil; Næs, Tormod; Skou, Peter B.; Solberg, Lars Erik; Dankel, Katinka R.; Basmoen, Silje A.; Wold, Jens Petter; Horn, Siri S.; Hillestad, Borghild; Poulsen, Nina A.; Christensen, Mette; Pieper, Theo; Afseth, Nils Kristian; Engelsen, Søren B.

I: Chemometrics and Intelligent Laboratory Systems, Bind 213, 104311, 2021.

Publikation: Bidrag til tidsskrift › Tidsskriftartikel › Forskning › fagfællebedømt

Harvard

Eskildsen, CE, Næs, T, Skou, PB, Solberg, LE, Dankel, KR, Basmoen, SA, Wold, JP, Horn, SS, Hillestad, B, Poulsen, NA, Christensen, M, Pieper, T, Afseth, NK & Engelsen, SB 2021, 'Cage of covariance in calibration modeling: Regressing multiple and strongly correlated response variables onto a low rank subspace of explanatory variables', Chemometrics and Intelligent Laboratory Systems, bind 213, 104311. https://doi.org/10.1016/j.chemolab.2021.104311

APA

Eskildsen, C. E., Næs, T., Skou, P. B., Solberg, L. E., Dankel, K. R., Basmoen, S. A., Wold, J. P., Horn, S. S., Hillestad, B., Poulsen, N. A., Christensen, M., Pieper, T., Afseth, N. K., & Engelsen, S. B. (2021). Cage of covariance in calibration modeling: Regressing multiple and strongly correlated response variables onto a low rank subspace of explanatory variables. Chemometrics and Intelligent Laboratory Systems, 213, [104311]. https://doi.org/10.1016/j.chemolab.2021.104311

Vancouver

Eskildsen CE, Næs T, Skou PB, Solberg LE, Dankel KR, Basmoen SA o.a. Cage of covariance in calibration modeling: Regressing multiple and strongly correlated response variables onto a low rank subspace of explanatory variables. Chemometrics and Intelligent Laboratory Systems. 2021;213. 104311. https://doi.org/10.1016/j.chemolab.2021.104311

Author

Eskildsen, Carl Emil ; Næs, Tormod ; Skou, Peter B. ; Solberg, Lars Erik ; Dankel, Katinka R. ; Basmoen, Silje A. ; Wold, Jens Petter ; Horn, Siri S. ; Hillestad, Borghild ; Poulsen, Nina A. ; Christensen, Mette ; Pieper, Theo ; Afseth, Nils Kristian ; Engelsen, Søren B. / Cage of covariance in calibration modeling : Regressing multiple and strongly correlated response variables onto a low rank subspace of explanatory variables. I: Chemometrics and Intelligent Laboratory Systems. 2021 ; Bind 213.

Bibtex

@article{4cd735cba9b4453384f6244c50aa1369,

title = "Cage of covariance in calibration modeling: Regressing multiple and strongly correlated response variables onto a low rank subspace of explanatory variables",

abstract = "In analytical chemistry, multivariate calibration is applied when substituting a time-consuming reference measurement (based on e.g. chromatography) with a high-throughput measurement (based on e.g. vibrational spectroscopy). An average error term, of the response variable, is often used to evaluate the performance of a calibration model. However, indirect relationships, between the response and explanatory variables, may be used for calibration. In such cases, model validity cannot necessarily be determined solely by the average error term. One should also consider the use of the models, as well as the validity of the indirect relationships in future samples. If the analyte of interest is partly quantified from signals of interfering compounds, then these interfering compounds will play a hidden role in the calibration. This hidden role may affect future use of the calibration model as strong covariance relationships between analyte estimates and interfering compounds may be imposed. Hence, such model cannot detect changes in the relationship between the analyte and interfering compounds. The problem is called the cage of covariance. This paper discusses the concept cage of covariance and possible consequences of applying models exposed to this issue.",

keywords = "Cage of covariance, Indirect models, Regression",

author = "Eskildsen, {Carl Emil} and Tormod N{\ae}s and Skou, {Peter B.} and Solberg, {Lars Erik} and Dankel, {Katinka R.} and Basmoen, {Silje A.} and Wold, {Jens Petter} and Horn, {Siri S.} and Borghild Hillestad and Poulsen, {Nina A.} and Mette Christensen and Theo Pieper and Afseth, {Nils Kristian} and Engelsen, {S{\o}ren B.}",

note = "Publisher Copyright: {\textcopyright} 2021 The Author(s)",

year = "2021",

doi = "10.1016/j.chemolab.2021.104311",

language = "English",

volume = "213",

journal = "Chemometrics and Intelligent Laboratory Systems",

issn = "0169-7439",

publisher = "Elsevier",

}

RIS

TY - JOUR

T1 - Cage of covariance in calibration modeling

T2 - Regressing multiple and strongly correlated response variables onto a low rank subspace of explanatory variables

AU - Eskildsen, Carl Emil

AU - Næs, Tormod

AU - Skou, Peter B.

AU - Solberg, Lars Erik

AU - Dankel, Katinka R.

AU - Basmoen, Silje A.

AU - Wold, Jens Petter

AU - Horn, Siri S.

AU - Hillestad, Borghild

AU - Poulsen, Nina A.

AU - Christensen, Mette

AU - Pieper, Theo

AU - Afseth, Nils Kristian

AU - Engelsen, Søren B.

PY - 2021

Y1 - 2021

N2 - In analytical chemistry, multivariate calibration is applied when substituting a time-consuming reference measurement (based on e.g. chromatography) with a high-throughput measurement (based on e.g. vibrational spectroscopy). An average error term, of the response variable, is often used to evaluate the performance of a calibration model. However, indirect relationships, between the response and explanatory variables, may be used for calibration. In such cases, model validity cannot necessarily be determined solely by the average error term. One should also consider the use of the models, as well as the validity of the indirect relationships in future samples. If the analyte of interest is partly quantified from signals of interfering compounds, then these interfering compounds will play a hidden role in the calibration. This hidden role may affect future use of the calibration model as strong covariance relationships between analyte estimates and interfering compounds may be imposed. Hence, such model cannot detect changes in the relationship between the analyte and interfering compounds. The problem is called the cage of covariance. This paper discusses the concept cage of covariance and possible consequences of applying models exposed to this issue.

AB - In analytical chemistry, multivariate calibration is applied when substituting a time-consuming reference measurement (based on e.g. chromatography) with a high-throughput measurement (based on e.g. vibrational spectroscopy). An average error term, of the response variable, is often used to evaluate the performance of a calibration model. However, indirect relationships, between the response and explanatory variables, may be used for calibration. In such cases, model validity cannot necessarily be determined solely by the average error term. One should also consider the use of the models, as well as the validity of the indirect relationships in future samples. If the analyte of interest is partly quantified from signals of interfering compounds, then these interfering compounds will play a hidden role in the calibration. This hidden role may affect future use of the calibration model as strong covariance relationships between analyte estimates and interfering compounds may be imposed. Hence, such model cannot detect changes in the relationship between the analyte and interfering compounds. The problem is called the cage of covariance. This paper discusses the concept cage of covariance and possible consequences of applying models exposed to this issue.

KW - Cage of covariance

KW - Indirect models

KW - Regression

U2 - 10.1016/j.chemolab.2021.104311

DO - 10.1016/j.chemolab.2021.104311

M3 - Journal article

AN - SCOPUS:85105252777

VL - 213

JO - Chemometrics and Intelligent Laboratory Systems

JF - Chemometrics and Intelligent Laboratory Systems

SN - 0169-7439

M1 - 104311

ER -

ID: 272016841