Scholarlink

The advent of machine learning techniques has paved the way for comprehensive athlete profiling, which is instrumental in identifying excellence in sports activities. The present investigation endeavours to investigate the correlation between Anthro-fitness variables and skateboarding performance. A total of 45 skateboarders were subjected to a comprehensive evaluation that included various skateboarding maneuvers, along with anthropometric and fitness assessments. The performances of the skateboarders were categorized, and a Random Forest (RF) classification model was developed to predict the grouping of the skateboarders. The cluster analysis revealed two distinct groups: High Anthro-fit Skateboarders (HAS) and Low Anthro-fit Skateboarders (LAS). The RF model demonstrated a high degree of accuracy, achieving 97% for training and 90% for testing. Further diagnostics of the model revealed that the HAS group exhibited superior performance in all the measured variables, except for abdominal circumference, compared to the LAS group. The integration of machine learning provides coaches and stakeholders with valuable insights, aiding in informed decision-making for athlete development and thereby enhancing competitiveness.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Profiling of High-Performance Skateboarders from Anthro-Fitness Variables: A Random Forest-Based-Machine Learning Analysis

  • 章节

  • 发表日期：

    2025年

查看全文
  • Rabiu Muazu Musa,
  • Aina Munirah Ab Rasid,
  • Che Nadia Che Samsudin,
  • Anwar P. P. Abdul Majeed,
  • Mohamad Razali Abdullah

摘要

The advent of machine learning techniques has paved the way for comprehensive athlete profiling, which is instrumental in identifying excellence in sports activities. The present investigation endeavours to investigate the correlation between Anthro-fitness variables and skateboarding performance. A total of 45 skateboarders were subjected to a comprehensive evaluation that included various skateboarding maneuvers, along with anthropometric and fitness assessments. The performances of the skateboarders were categorized, and a Random Forest (RF) classification model was developed to predict the grouping of the skateboarders. The cluster analysis revealed two distinct groups: High Anthro-fit Skateboarders (HAS) and Low Anthro-fit Skateboarders (LAS). The RF model demonstrated a high degree of accuracy, achieving 97% for training and 90% for testing. Further diagnostics of the model revealed that the HAS group exhibited superior performance in all the measured variables, except for abdominal circumference, compared to the LAS group. The integration of machine learning provides coaches and stakeholders with valuable insights, aiding in informed decision-making for athlete development and thereby enhancing competitiveness.