Maximal Oxygen Uptake (VO2max) is a crucial parameter that indicates the maximum oxygen usage capacity of the body during exercise. VO2max is determined by multiple factors such as age, genetics, gender and training status and can be improved by regular aerobic exercise. Knowledge of an athlete's VO2max assists coaches in designing personalized training programs to enhance their endurance and aerobic fitness, resulting in improved athletic performance.

However, assessing VO2max accurately requires a maximal exercise test involving direct measurement of oxygen consumption, which can be invasive, time-consuming, and costly. Several indirect methods have been developed to estimate VO2max, including Heart Rate Variability (HRV) analysis, which is a non-invasive, quick, and cost-effective alternative to direct measurement. Wearable Heart Rate (HR) monitors are widely used in sports to provide physiological insights into athletes' well-being and performance. These monitors are unobtrusive, easy to use, and provide reliable HR measurements that enable the estimation of athletes' cardiorespiratory fitness, quantified by their VO2max.

In this study, HRV features were extracted from both exercise and recovery segments to estimate the VO2max of 856 athletes who underwent Graded Exercise Testing. Three different machine learning models were used to estimate VO2max, and three feature selection methods were applied to avoid overfitting of the models and obtain relevant features. This experimental approach validates the utility of HRV to estimate VO2max in a large population of athletes. Additionally, it contributes to the usefulness of wearable HR monitors for assessing cardiorespiratory fitness among athletes, as it provides a non-invasive and cost-effective method for estimating VO2max. Overall, this research work contributes to the growing body of research supporting the use of wearable HR monitors for assessing athletes' cardiorespiratory fitness.