Big Data powered recommendations for recreational athletes

Investigators: Barry Smyth, Brian Caulfield, Aonghus Lalor

A group led by Barry Smyth has been collaborating a leading fitness app to better understand how recreational athletes exercise, train, and compete. This research has been enabled through an application partner making their entire dataset (up to 2017) available to Insight for research purposes.  Studies to date have focused on marathon runners to answer questions such as:

• how should runners (new or experienced) train to ensure that they can complete the marathon distance and/or achieve their personal goal-times?(Doherty 2019);
• is it possible to predict whether runners will become injured based on their training patterns; can we predict a runner’s likely finish-time based on their training and/or recent race performances?(Smyth 2020);
• can we help runners to plan their race so that they maximise their performance and avoid hitting the wall? (Smyth 2021)(Doherty 2020);
• is it possible to advise runners on how best to adjust their pacing, in real-time, during a race? (Berndsen 2020)(Berndsen 2019)

All of these are important questions to answer especially given the time and effort that people devote to their training. The work is possible because of the collaboration that exists in UCD between sports scientists and machine learning researchers.

Read about Sarah Dillon’s work on running injuries