Bio-inspired algorithms for the characterization of excellent performance in handball players: A data-driven methodology

Bio-inspired algorithms have been successfully applied to solve complex optimization problems. They are also widely used to train and optimize machine learning and data-driven models, providing competitive results. This study presents a novel data-driven approach to identify and quantify the factors...

Descripción completa

Detalles Bibliográficos
Autores: López Gómez, Julio Alberto, Romero Francisco P., Angulo, Eusebio
Tipo de recurso: artículo
Fecha de publicación:2025
País:España
Institución:Universidad de Castilla-La Mancha
Repositorio:RUIdeRA. Repositorio Institucional de la UCLM
OAI Identifier:oai:ruidera.uclm.es:10578/47686
Acceso en línea:https://doi.org/10.1016/j.eswa.2025.126821
https://hdl.handle.net/10578/47686
Access Level:acceso abierto
Palabra clave:Sports analytics
Handball performance
Bio-inspired algorithms
Data-driven modeling
Optimization in sports
Descripción
Sumario:Bio-inspired algorithms have been successfully applied to solve complex optimization problems. They are also widely used to train and optimize machine learning and data-driven models, providing competitive results. This study presents a novel data-driven approach to identify and quantify the factors that characterize the excellent performance of handball players, depending on the specific position in which they play. This will give us the most important characteristics that differentiate the most excellent players in their positions. Based on bio-inspired algorithms, this research delves into the complex optimization problems inherent in sports analytics. The study utilizes data from the Women’s European Handball Championship, employing seven distinct algorithms, of which six are different bio-inspired algorithms - including a hybrid bio-inspired algorithm - and one Consensus-Based Aggregation algorithm to analyze and assign weights to each player’s actions during a match. This approach is further validated by comparing the findings against the top five players in each position as recognized by the European Handball Federation (EHF). Subsequently, the established model’s robustness and applicability are tested using data from the Women’s World Handball Championships.