Communications of the ACM
Novel Algorithm Proposed for Selecting Variables Efficiently
The algorithm combines six weight vectors and uses a threshold search strategy to seek the optimal weight vector to extract useful information from the spectrum.
Credit: Hefei Institutes of Physical Science/Chinese Academy of Sciences
Researchers at the Chinese Academy of Sciences' Hefei Institutes of Physical Science developed a variable selection algorithm for chemometrics applications.
The multi-weight vector optimal selection and bootstrapping soft shrinkage (MWO-BOSS) algorithm aims to make more efficient the process of identifying an optimal wavelength combination when developing spectral prediction models.
MWO-BOSS selects the optimal weight vector from among six weight vectors (selectivity ratio, variable importance in projection, frequency vector, reciprocal of residual variance vector, regression coefficient, and significance multivariate correlation) to extract valuable information from the spectrum by employing a threshold search strategy.
In tests on publicly available datasets, the algorithm was successful in choosing variables efficiently and enhancing the predictive ability of the model.
From Chinese Academy of Sciences
View Full Article
Abstracts Copyright © 2023 SmithBucklin, Washington, D.C., USA
No entries found