Topological Data Analysis Enhances Tennis Match Prediction Accuracy

Researchers have developed novel methods for predicting tennis match outcomes by applying topological data analysis and graph theory to ATP singles match data from 2000-2025. One approach uses lower-star filtration on player competitive networks, extracting topological features through persistent homology. This method, combined with graph-theoretic and ranking features, achieved 66.2% accuracy with a Random Forest model, indicating that topological features provide a complementary signal to traditional metrics. Notably, a topology-only model maintained 63.56% accuracy, demonstrating the power of network-derived features alone. AI

RANK_REASON The cluster contains a research paper detailing novel methods for sports analytics. [lever_c_demoted from research: ic=1 ai=0.7]

Read on arXiv stat.ML →

• ATP singles matches
• Graph Theory
• lower-star filtration
• Modified Katz similarity index
• persistent homology
• Topological Data Analysis

• paper
• other

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →