CHOOSING AN EFFICIENT ALGORITHM FOR SOLVING THE CLASSIFICATION PROBLEM

In this work, the selection of an effective algorithm for solving the classification problem was considered. The general mathematical formulation of the problem of classification is given and the available algorithms for solving it are analyzed. In addition, a taxonomy of the most used classification algorithms based on their ideological origins is presented. Classification errors were compared using K-Fold Cross-Validation testing. The coefficient of error in classification was taken into account as an efficiency indicator. The obtained results were checked on the basis of several data sets and the reliability of the algorithm was compared.