A novel variable selection method based on combined moving window and intelligent optimization algorithm for variable selection in chemical modeling

Pengfei Zhang; Zhuopin Xu; Qi Wang; Shuang Fan; Weimin Cheng; Haiping Wang; Yuejin Wu

doi:10.1016/j.saa.2020.118986

A novel variable selection method based on combined moving window and intelligent optimization algorithm for variable selection in chemical modeling

Spectrochim Acta A Mol Biomol Spectrosc. 2021 Feb 5:246:118986. doi: 10.1016/j.saa.2020.118986. Epub 2020 Sep 25.

Authors

Pengfei Zhang¹, Zhuopin Xu², Qi Wang³, Shuang Fan², Weimin Cheng², Haiping Wang², Yuejin Wu⁴

Affiliations

¹ Hefei Institutes of Physical Science, Chinese Academy of Sciences, Hefei 230031, China.
² Hefei Institutes of Physical Science, Chinese Academy of Sciences, Hefei 230031, China; University of Science and Technology of China, Hefei 230026, China.
³ Hefei Institutes of Physical Science, Chinese Academy of Sciences, Hefei 230031, China. Electronic address: wangqi@ipp.ac.cn.
⁴ Hefei Institutes of Physical Science, Chinese Academy of Sciences, Hefei 230031, China. Electronic address: yjwu@ipp.ac.cn.

PMID: 33032116
DOI: 10.1016/j.saa.2020.118986

Abstract

We propose a new wavelength selection algorithm based on combined moving window (CMW) and variable dimension particle swarm optimization (VDPSO) algorithm. CMW retains the advantages of the moving window algorithm, and different windows can overlap each other to realize automatic optimization of spectral interval width and number. VDPSO algorithms improve the PSO algorithm. They can search the data space in different dimensions, and reduce the risk of limited local extrema and over fitting. Four different high-performance variable selection algorithms-BOSS, VCPA, iVISSA and IRF-are compared in three NIR data sets (corn, beer and fuel). The results show that VDPSO-CMW has better performance. The Matlab codes for implementing PSO-CWM and VDPSO-CMW are freely available on the website: https://www.mathworks.com/matlabcentral/fileexchange/75828-a-variable-selection-method.

Keywords: Intelligent optimization algorithm; Multivariate calibration; Near-infrared spectroscopy; Particle swarm optimization; Variable selection.