BUILDING A PREDICTIVE CLASSIFICATION MODEL USING ARTIFICIAL NEURAL NETWORK TECHNIQUES AND IN TERMS OF ANTHROPOMETRIC MEASUREMENTS, BIO-KINETIC ABILITIES, AND SKILL PERFORMANCE OF YOUNG BOXERS

Boxing is a combat sport that requires a high level of physical and bio-kinetic abilities due to its nature, which is characterized by rapid reaction, accuracy, balance, and responsiveness to changing competitive situations. The importance of this sport lies in the need for precise scientific tools that contribute to enhancing selection and classification processes, which positively affects performance effectiveness and guides the training process scientifically. The research problem was represented by the absence of accurate objective models that rely on the analysis of body measurements and bio-kinetic abilities as a basis for classifying junior boxers. This weakens training outcomes and negatively affects skill achievement. Hence, the importance of the research lies in contributing to the construction of a predictive classification model for skill level using artificial neural network techniques and the significance of body measurements and bio-kinetic abilities in junior boxers. The research aimed to identify the most prominent body and bio-kinetic variables that distinguish players with a high skill level, create a classification model that helps predict performance levels, and provide a database that supports coaches in designing effective training programs tailored to the characteristics of players. The researcher adopted the descriptive survey approach, given its suitability to the nature of the problem. Both the exploration and main experiments were conducted in Baghdad Governorate, inside the Al-Ittihad Sports Club hall. The research sample included (70) junior players representing the Al-Ittihad, Al-Arabi, and Al-Hussein clubs, for the 2022-2023 sports season. They underwent physical and bio-kinetic tests that measured multiple aspects related to skill performance in boxing. The results showed clear individual differences among the players in their bio-kinetic abilities, which was reflected in their skill performance. The model, based on artificial neural network techniques, also proved highly effective in accurately classifying players, enhancing its potential for use in selection and guidance. According to the results, the researcher recommends the adoption of artificial intelligence techniques, particularly artificial neural networks, in the evaluation and selection processes of athletes, while emphasizing the need to organize training workshops to qualify training personnel in this field. This will contribute to improving performance and raising the competitive levels of junior boxing players.