Investigation of machine learning classification methods effectiveness

The question of data collection and processing constantly arises in our lives. Therefore, to use this data, people need to know how to work with it properly. The paper's primary purpose is to investigate machine learning classification methods and compare their effectiveness and accuracy to solve the classification problem. The authors solved the classification problem of a company's bankruptcy with a given economic and financial features. For this, a dataset was used, based on which the efficiency and the quality of application of several existing classification algorithms are investigated. In particular, the following classifiers were used: ordinary and linear methods of Support Vector Machines, Extra Trees, Random Forest, Decision Tree, Logistic Regression, MLP classifier, Gradient Boosting, Naive Bayes Classifier. Estimates of accuracy, precision, recall, F1-measure, ROC are given to evaluate the classification quality and choose the best classifier.