Using variable combination population analysis for variable selection in multivariate calibration

• Published in: Analytica chimica acta Vol. 862; pp. 14 - 23
• Main Authors: Yun, Yong-Huan, Wang, Wei-Ting, Deng, Bai-Chuan, Lai, Guang-Bi, Liu, Xin-bo, Ren, Da-Bing, Liang, Yi-Zeng, Fan, Wei, Xu, Qing-Song
• Format: Journal Article
• Language: English
• Published: Netherlands Elsevier B.V 03.03.2015
• Subjects: Algorithms; analytical chemistry; Calibration; data collection; evolution; Exponentially decreasing function; Internet; least squares; Least-Squares Analysis; Model population analysis; Models, Statistical; Monte Carlo Method; Multivariate Analysis; Multivariate calibration; Partial least squares; selection methods; Variable combination; Variable selection; wavelengths; Variable combination; Exponentially decreasing function; Model population analysis; Partial least squares; Multivariate calibration; Variable selection
• ISSN: 0003-2670; 1873-4324; 1873-4324
• DOI: 10.1016/j.aca.2014.12.048

[Display omitted] •Uses a novel sampling strategy that considers the interaction effect among variables.•Utilizes model population analysis to explore the useful information from a population of models.•Employs exponentially decreasing function to iteratively filter the variables and shrink the variable space.•Performs well when compared with CARS, GA–PLS, MC-UVE-PLS and IRIV. Variable (wavelength or feature) selection techniques have become a critical step for the analysis of datasets with high number of variables and relatively few samples. In this study, a novel variable selection strategy, variable combination population analysis (VCPA), was proposed. This strategy consists of two crucial procedures. First, the exponentially decreasing function (EDF), which is the simple and effective principle of ‘survival of the fittest’ from Darwin’s natural evolution theory, is employed to determine the number of variables to keep and continuously shrink the variable space. Second, in each EDF run, binary matrix sampling (BMS) strategy that gives each variable the same chance to be selected and generates different variable combinations, is used to produce a population of subsets to construct a population of sub-models. Then, model population analysis (MPA) is employed to find the variable subsets with the lower root mean squares error of cross validation (RMSECV). The frequency of each variable appearing in the best 10% sub-models is computed. The higher the frequency is, the more important the variable is. The performance of the proposed procedure was investigated using three real NIR datasets. The results indicate that VCPA is a good variable selection strategy when compared with four high performing variable selection methods: genetic algorithm–partial least squares (GA–PLS), Monte Carlo uninformative variable elimination by PLS (MC-UVE-PLS), competitive adaptive reweighted sampling (CARS) and iteratively retains informative variables (IRIV). The MATLAB source code of VCPA is available for academic research on the website: http://www.mathworks.com/matlabcentral/fileexchange/authors/498750.