Дослідження роботи алгоритму gradient boosting для класифікації транспорту

Abstract

This study explores the application of the Gradient Boosting (GB) algorithm in classifying transport images, focusing on its performance and limitations. As an ensemble learning method, GB enhances classification accuracy through the sequential combination of multiple decision trees. The findings reveal that GB achieves an overall accuracy of 57%, with varying F1-scores across different transport categories. Although GB does not deliver near-perfect multi-class classification, it remains a viable approach for analyzing transport images within a reasonable range of accuracy.