Powerlifting score prediction using a machine learning method.

Area of Science:

• Sports Science
• Artificial Intelligence
• Machine Learning

Background:

• Predicting athletic performance is crucial in sports.
• Accurate powerlifting score prediction can inform training and competition strategies.
• Existing methods may lack the precision required for reliable forecasting.

Purpose of the Study:

• To develop an accurate artificial intelligence model for predicting powerlifting scores.
• To enhance prediction accuracy by optimizing the reservoir computing extreme learning machine.
• To provide a reliable tool for experts to estimate powerlifting results before events.

Main Methods:

• Collected powerlifter characteristics data.
• Utilized reservoir computing extreme learning machine for initial predictive modeling.
• Implemented the whale optimization algorithm to optimize the reservoir computing extreme learning machine parameters.

Main Results:

• Achieved a coefficient of determination (R²) of 0.7958, indicating strong model fit.
• Obtained a root-mean-square error of prediction (RMSE) of 16.73, demonstrating prediction accuracy.
• The optimized model significantly improved upon baseline prediction capabilities.

Conclusions:

• The proposed AI method, combining reservoir computing and whale optimization, effectively predicts powerlifting scores.
• This approach offers a reliable basis for pre-competition performance evaluation.
• Further research can explore broader applications of this optimized AI technique in sports analytics.