COMPARISON OF BINARY CLASSIFICATION METRICS FOR EVALUATING THE EFFECTIVENESS OF MACHINE LEARNING ALGORITHMS

Shirokova Ekaterina Vasilyevna

The article compares various binary classification metrics to evaluate the effectiveness of machine learning algorithms. Various approaches are considered, such as the support vector machine method, the decision tree method, the random forest method, and the gradient boosting method. The advantages and disadvantages of metrics in the case of balanced and unbalanced data are analyzed, as well as examples of their implementation using the Python Scikit-learn library.

Keywords:machine learning, data classification, class imbalance.