Quantitative Study of Tennis Match Momentum and Flow Based on Machine Learning

Dishen Yang; Qingyang Zhang; Baihe Luan

doi:10.62051/56npsh86

Authors

Dishen Yang
Qingyang Zhang
Baihe Luan

DOI:

https://doi.org/10.62051/56npsh86

Keywords:

Levenberg-Marquardt Back Propagation, Neural Network, Quantitative Model, Big Data.

Abstract

Momentum is a critical factor influencing the dynamics and outcomes of tennis matches. To enhance the predictive accuracy of these fluctuations, this study utilizes machine learning techniques and big data analytics to improve the prediction of these fluctuations. The study conducts a comprehensive analysis to identify the correlation between a player's momentum and 14 key features in a tennis match. An optimized BP neural network model, based on Levenberg-Marquardt theory, was developed to predict match flow and quantify the stalemate degree. The model is evaluated using a confusion matrix and ROC curve, affirming its predictive validity, where the results revealed an F1 Score and an AUC, both exceeding 0.5. With big data, this approach not only enhances the spectator experience by visualizing match dynamics but also aids in strategy development and training optimization for competitors. This research highlights the practical applications of quantitative modeling in understanding and forecasting the pivotal moments in tennis.

Downloads

Download data is not yet available.

References

[1] [1] DEPKEN C A, GANDAR J M, SHAPIRO D A. Set-level Strategic and Psychological Momentum in Best-of-three-set Professional Tennis Matches[J/OL]. Journal of Sports Economics, 2022, 23(5): 598-623.

[2] [2] RUBIO J de J. Stability Analysis of the Modified Levenberg–Marquardt Algorithm for the Artificial Neural Network Training[J/OL]. IEEE Transactions on Neural Networks and Learning Systems, 2021, 32(8): 3510-3524.

[3] [3] HUANG K, LI J. Predictive control of seismic responses of a novel seismic isolation – non-seismic isolation composite structure system based on Levenberg-Marquardt algorithm[J/OL]. Structures, 2023, 53: 373-381.

[4] [4] DU X. Prediction of Power Consumption of Hydroelectric Power Station by Levenberg-Marquardt-BP Algorithm[C/OL]//2021 2nd International Seminar on Artificial Intelligence, Networking and Information Technology (AINIT). 2021: 43-46[2024-06-02].

[5] [5] GUAN S, WANG X. Optimization analysis of football match prediction model based on neural network[J/OL]. Neural Computing and Applications, 2022, 34(4): 2525-2541.

[6] [6] ÜNAL T. Predicting tennis match outcome: a machine learning approach using the SRP-CRISP-DM framework[D/OL]. Middle East Technical University, 2023[2024-06-07].

[7] [7] HOO Z H, CANDLISH J, TEARE M D. What is an ROC curve?[J]. Emergency Medicine Journal, 2017, 34(6): 357-359.

[8] [8] MELONIO V P de F, AOKI M S, ARRUDA A F S, et al. Analysis of serve and serve return on different surfaces in elite tennis players[J/OL]. Revista Brasileira de Cineantropometria & Desempenho Humano, 2021, 23: e76603.

[9] [9] OLIVEIRA V, MENAYO R, FUENTES-GARCÍA J P. Training Tennis through Induced Variability and Specific Practice: Effects on Performance in the Forehand Approach Shot[J/OL]. Applied Sciences, 2024, 14(8): 3287.

[10] [10] JIAN-TAO S. Evaluation and Analysis of an Industrial Cluster Based on the BP Neural Network and LM Algorithm[J/OL]. Wireless Communications & Mobile Computing (Online), 2022, 2022[2024-06-07].

[11] [11] MUSCHELLI J. ROC and AUC with a Binary Predictor: a Potentially Misleading Metric[J/OL]. Journal of Classification, 2020, 37(3): 696-708.