### Abstract

Additional benefits arise from the use of LASSO and ENET. While LASSO uses only ℓ1 norm on regression coefficients, ENET combines the best of both worlds by using ℓ1 and ℓ2 norms. The former (ℓ1) provides variable selection while the latter (ℓ2) handles collinearity via shrinkage of regression coefficients. Rational functions are highly collinear if full rank is used and, thus, not necessarily robust unless either ℓ1 or ℓ2 norm is used on the regression coefficients. The use of ℓ1 norm allows for a more parsimonious model that can potentially be more robust. This is contrary to the use of a broadband spectrum that is likely to be contaminated at some point in the future by unknown spectral interferences. The real benefits seem to originate from the combination of rational functions and ENET. Note that LASSO solutions form a subset of ENET solutions and are thus included in ENET.

Original language | English |
---|---|

Pages (from-to) | 57-68 |

Number of pages | 12 |

Journal | Analytica Chimica Acta |

Volume | 768 |

DOIs | |

Publication status | Published - 2013 |

MoE publication type | A1 Journal article-refereed |

Event | XIII Conference on Chemometrics in Analytical Chemistry - Budapest, Hungary Duration: 25 Jun 2012 → 29 Jun 2012 |

### Fingerprint

### Keywords

- chemometrics
- elastic net
- LASSO
- multi-point
- multivariate calibration
- NIR
- rational function
- RF-ENET

### Cite this

*Analytica Chimica Acta*,

*768*, 57-68. https://doi.org/10.1016/j.aca.2013.01.005

}

*Analytica Chimica Acta*, vol. 768, pp. 57-68. https://doi.org/10.1016/j.aca.2013.01.005

**Parsimonious and robust multivariate calibration with rational function Least Absolute Shrinkage and Selection Operator and rational function Elastic Net.** / Teppola, Pekka (Corresponding Author); Taavitsainen, V.-M.

Research output: Contribution to journal › Article › Scientific › peer-review

TY - JOUR

T1 - Parsimonious and robust multivariate calibration with rational function Least Absolute Shrinkage and Selection Operator and rational function Elastic Net

AU - Teppola, Pekka

AU - Taavitsainen, V.-M.

PY - 2013

Y1 - 2013

N2 - This paper presents new methods for multivariate calibration. A unique aspect is that this approach uses rational functions with either Least Absolute Shrinkage and Selection Operator (LASSO) or Elastic Net (ENET), and builds parsimonious models in an automated way via cross-validation. Rational function modeling provides robustness, as will be briefly demonstrated. Interestingly, rational function models are also flexible, in that occasionally they are reduced to ordinary linear models based on cross-validation. Thus, model complexity is not forced to take the form of rational functions.Additional benefits arise from the use of LASSO and ENET. While LASSO uses only ℓ1 norm on regression coefficients, ENET combines the best of both worlds by using ℓ1 and ℓ2 norms. The former (ℓ1) provides variable selection while the latter (ℓ2) handles collinearity via shrinkage of regression coefficients. Rational functions are highly collinear if full rank is used and, thus, not necessarily robust unless either ℓ1 or ℓ2 norm is used on the regression coefficients. The use of ℓ1 norm allows for a more parsimonious model that can potentially be more robust. This is contrary to the use of a broadband spectrum that is likely to be contaminated at some point in the future by unknown spectral interferences. The real benefits seem to originate from the combination of rational functions and ENET. Note that LASSO solutions form a subset of ENET solutions and are thus included in ENET.

AB - This paper presents new methods for multivariate calibration. A unique aspect is that this approach uses rational functions with either Least Absolute Shrinkage and Selection Operator (LASSO) or Elastic Net (ENET), and builds parsimonious models in an automated way via cross-validation. Rational function modeling provides robustness, as will be briefly demonstrated. Interestingly, rational function models are also flexible, in that occasionally they are reduced to ordinary linear models based on cross-validation. Thus, model complexity is not forced to take the form of rational functions.Additional benefits arise from the use of LASSO and ENET. While LASSO uses only ℓ1 norm on regression coefficients, ENET combines the best of both worlds by using ℓ1 and ℓ2 norms. The former (ℓ1) provides variable selection while the latter (ℓ2) handles collinearity via shrinkage of regression coefficients. Rational functions are highly collinear if full rank is used and, thus, not necessarily robust unless either ℓ1 or ℓ2 norm is used on the regression coefficients. The use of ℓ1 norm allows for a more parsimonious model that can potentially be more robust. This is contrary to the use of a broadband spectrum that is likely to be contaminated at some point in the future by unknown spectral interferences. The real benefits seem to originate from the combination of rational functions and ENET. Note that LASSO solutions form a subset of ENET solutions and are thus included in ENET.

KW - chemometrics

KW - elastic net

KW - LASSO

KW - multi-point

KW - multivariate calibration

KW - NIR

KW - rational function

KW - RF-ENET

U2 - 10.1016/j.aca.2013.01.005

DO - 10.1016/j.aca.2013.01.005

M3 - Article

VL - 768

SP - 57

EP - 68

JO - Analytica Chimica Acta

JF - Analytica Chimica Acta

SN - 0003-2670

ER -