Prediction the number of coaching certificates in Türkiye with machine learning: a comparative study of artificial neural networks and adaptive neuro-fuzzy inference system

Abstract

This study aims to quantify the issuance of coaching certifications in Türkiye using artificial neural networks (ANN) and an adaptive neuro-fuzzy inference system (ANFIS). Coaching plays a pivotal part in athlete development, and meticulous preparation is essential for the enduring success of sports. The study compared the performance of ANN and ANFIS models using coaching certificate data from 2002 to 2022. The findings indicate that the ANFIS model yields more precise predictions compared to the ANN model, with R² = 0.9998 and MAPE = 0.9. The ANN model has somewhat inferior performance, with R² = 0.9420 and MAPE = 8.3. In summary, the ANFIS model has demonstrated superior accuracy in predicting non-linear data sets and serves as an excellent instrument for determining the requisite number of coaches. This study's findings indicate that artificial intelligence-based models for projecting coaching certificate numbers in Türkiye are beneficial and serve as a valuable asset for strategic planning in sport management.