A Novel Binary QUasi-Affine TRansformation Evolution (QUATRE) Algorithm and Its Application for Feature Selection

QUasi-Affine TRansformation Evolution (QUATRE) algorithm is a new intelligent computing algorithm based on matrix iteration behavior. Binary QUATRE (BQUATRE) is a binary version that can be used to solve binary application problems. From continuous to binary arithmetic is a crucial part of the binary version. In order to convert the continuous type to the binary type, this paper proposes a simple and effective conversion method. After the benchmark function test, it proves that the improved binary QUATRE method has strong competitiveness. Finally, the feature selection problem can be successfully solved in the UCI data set, and a higher classification accuracy can be obtained with a smaller number of features.