Evaluation of a Training Recommendation Strategy for Trail Runners.

Finding the optimal training, i.e. the training that enables the athlete to achieve the best performance while preserving their physical and psychological integrity, is a multi-factorial issue.

Each individual's background and training capacity are very different, hence the need for personalised training. The availability of data combined with advances in Artificial Intelligence (data modelling and analysis sciences, big data, machine learning, deep learning, etc.) offer the opportunity to refine our understanding of training and adapt recommendations according to the runner's profile for a given objective (achieving a time over a distance, completing an event, etc.). The strategy that the investigators wish to evaluate as part of this trial could make it possible to recommend an optimal training load, as well as the distribution of intensities (as a percentage of the aerobic threshold, anaerobic threshold, VVO2max, etc). The two features are (i) the personalisation will be based on training history, the level of the runner, age and gender, and the runner can focus on training, and (ii) the strategy will be adaptive, i.e. the algorithm will update the training load in real time based on contextual data about the athlete (level of fatigue via heart rate variability, feedback on recent training sessions, feelings of stress).