How Can Machine Learning Determine Whether a Women's Tennis Player Will Make it to Top 100?

Abstract

There are a  lot of speculations within and outside of the tennis community about whether factors like height, age, and nationality play a role in the success of a tennis player. For this study, ‘success’ is defined as making it to the Top 100 ranked list. There have been studies in the past associating height of a tennis player with success, but this has primarily been done for men’s tennis players. In this study, we not only establish the relation between height and success of women tennis players but also consider two additional factors: age and nationality. We also mathematically conclude using Pearson’s correlation coefficient whether there is any statistical correlation between these three factors and success. Once we establish the relationship, we develop an AI model to predict future successful players based on historical tennis data. Since some of the earlier studies have already considered height as one of the success factors, our machine learning model uses Naïve Bayes’ to determine the probability of success using all three factors to predict success with an accuracy of 0.67 for dataset used. The individual Pearson correlation coefficients for height and age with success, demonstrating the applicability of factors in identifying a player’s potential for success are 0.23 and 0.19 respectively. Further research can be conducted by using more factors or larger dataset and could foster greater understanding of female success in tennis.

Keywords: Age, AI, Bayes’ Theorem, Factors, Height, Machine learning, Naïve Bayes, Nationality, Pearson’s Correlation Coefficient, Ranking, Tennis, Top 100