Basketball Data Analysis Using Spark Framework and K-Means Algorithm

Area of Science:

• Sports Science and Analytics
• Data Science and Big Data
• Computational Intelligence

Background:

• Wearable and sensing technologies generate vast amounts of sports data, creating challenges for traditional analysis.
• Big data analytics are crucial for improving the effectiveness of basketball training and player evaluation.
• Assessing defensive performance under high-pressure situations is vital for team strategy.

Purpose of the Study:

• To propose an efficient big data processing framework for basketball training analysis.
• To enhance player performance assessment using advanced algorithms and distributed computing.
• To provide tools for recruiters and trainers to evaluate and improve team performance.

Main Methods:

• Implementation of the Spark framework for in-memory big data processing.
• Utilization of a cuckoo search algorithm for swarm intelligence optimization.
• Application of K-clustering algorithm within a Spark distributed environment for enhanced analysis.
• Examination of defensive performance metrics in high-pressure game scenarios.

Main Results:

• The proposed approach demonstrates superior clustering performance and practical utility compared to existing methods.
• The framework effectively analyzes player defensive performance under pressure.
• Significant improvements were observed in assessing the impact of moving and shooting training.

Conclusions:

• The developed big data framework offers a powerful tool for basketball analytics.
• This method aids in understanding player qualities and optimizing team strategies.
• The findings support enhanced training effectiveness, particularly in shooting and defensive skills.