Preprocessing COVID-19 radiographic images by evolutionary column subset selection

Abstract

Column subset selection is a hard combinatorial optimization problem with applications in operations research, data analysis, and machine learning. It involves the search for fixed–length subsets of columns from large data matrices and can be used for low–rank approximation of high–dimensional data. It can be also used to preprocess data for image classification. In this work, we study column subset selection in the context of radiography image analysis and concentrate on the detection of COVID-19 from chest X–ray imagery.