RESEARCH PAPERAdaptive Variable Re-weighting and Shrinking Approach for Variable Selection in Multivariate Calibration for Near-infrared Spectroscopy
Graphical abstract
A new variable selection method, named AVRSA, is proposed. It uses weighted bootstrap sampling to generate a random sub-model group, then selects sub-models, whose performance is better than the average performance of the optimal sub-model group generated in the last iteration to form a new optimal sub-model group, and finally updates the variable weights and shrinks the variable space by evaluating the optimal sub-model group. Repeat until no optimal sub-model is generated, then the optimal informative variables are obtained.
References (42)
- et al.
Measurement
(2020) - et al.
Carbohydr. Polym.
(2019) - et al.
Food Chem.
(2018) - et al.
TrAC-Trends Anal. Chem.
(2016) - et al.
Chemom. Intell. Lab. Syst.
(2001) - et al.
Chemom. Intell. Lab. Syst.
(2012) - et al.
Anal. Chim. Acta
(2009) - et al.
Chemom. Intell. Lab. Syst.
(2008) - et al.
Anal. Chim. Acta
(2014) - et al.
Anal. Chim. Acta
(2019)
Anal. Chim. Acta
Anal. Chim. Acta
Anal. Chim. Acta
Anal. Chim. Acta
Chin. J. Anal. Chem.
Chemom. Intell. Lab. Syst.
Chemom. Intell. Lab. Syst.
Chemom. Intell. Lab. Syst.
TrAC-Trends Anal. Chem.
J. Braz. Chem. Soc.
Sustainability
Cited by (0)
This work was supported by the National Natural Science Foundation of China (No. 31871571), the Key Technologies R&D Program of Shanxi Province, China (Nos. 201903D211002, 201603D221037-3), the Shanxi Province Applied Basic Research Project of China (No. 201801D221299), and the Science and Technology Innovation Fund of Shanxi Agricultural University, China (Nos. 201308, 2020BQ32).