Problem Statement and Approach: In basketball, the success of free throws is important in terms of the ball's exit angle, its proper position in the air and its optimum speed kinematic features. This study investigates the kinematic characteristics of basketball players executing free throws both before and after experiencing fatigue, using artificial Intelligence (AI). Materials and Methods: We used various supervised machine learning algorithms including: k-Nearest Neighbours (k-NN), Naïve Bayes, Support Vector Machines (SVM), Artificial Neural Networks (ANN), Linear Discriminant Analysis (LDA), and Decision Tree. These algorithms were used for classifying features derived from motion data collected from players to uncover patterns and variations in their shooting mechanics under different levels of fatigue. The elbow, trunk, knee and ankle joint angles are measured at the ball release point when the players make successful and unsuccessful shots before and after fatigue. There are two used methods for classification of these features: the first one is used directly row data; the other is used reduced data by using principal component analysis (PCA). For both, data normalized between 0-1 before applying to the classifier. Results: We obtained the best classification accuracy as 98.44% by using Naïve Bayes classifier for row data. Moreover, the results of ANN with reduction data by using PCA have shown the best classification accuracy as 95.31%. The findings reveal distinct patterns and variations in shooting mechanics induced by fatigue and highlight the effectiveness of machine learning models in analysing biomechanical data. Discussion and conclusion: These results can aid in developing training programs to enhance performance and consistency in fatigued states. This research underscores the potential of AI and data-driven approaches in sports biomechanics to uncover valuable insights into athlete performance and fatigue management.
© Copyright 2024 Journal of Physical Education and Sport. University of Pitesti. Alle Rechte vorbehalten.