Fencing is an Olympic sport with three modalities differentiated by the weapon used: épée, foil, and sabre. Each involves different physical demands due to the characteristics of the weapon, such as its weight and the specific combat technique required. The objective of this study was to develop a predictive model to determine the op- timal fencing modality based on the athletes' anthropometric characteristics and physical capacities. The study was quantitative, cross-sectional, and observational, and was conducted with active fencers from the Valle del Cauca league in Colombia. Anthropometric (weight, wingspan, muscle mass, and fat percentage) and physical variables (measured using the McCloy battery) were evaluated, along with additional data such as age and VO2 max. To improve the representativeness of minority classes, the SMOTE oversampling technique was applied. Five classification models (LDA, KNN, CART, NB, and SVM) were tested, the Decision Tree being the one that obtained the best performance after being optimized with Grid Search. The results showed that the percentage of muscle and body fat, as well as the squat with jump, significantly influence the identification of the weapon used, allow- ing for more specific and efficient training strategies according to the competition modality.
© Copyright 2025 SportMont. Montenegrin Sports Academy. Alle Rechte vorbehalten.