A 3-year strategic partnership on sports data science between the Centre for Data Science (CDS), the Australian Institute of Sport (AIS) and the Queensland Academy of Sport (QAS) was launched in the past few months. With a drive towards data informed decision making across the high performance sports network nationally, a number of collaborative, interdisciplinary research and scholarship opportunities ranging from VRES, to honours, masters and PhD have developed.
Students will be working with data scientists in the centre, real-world sports data (e.g. from wearable sensors, video analytics and more), and partner sports organisations. The focus of the partnership is to push the state of the art in sports data science to better support coaches, athletes and sports scientists in achieving greater performance and health (e.g. minimise injuries) outcomes.
Potential topics will involve applying and/or developing statistical and machine learning methods to:
- Predict and benchmark athlete seasonal progression given training, lab test, competition and other data,
- Perform sensor fusion of accelerometer and video tracking data,
- Predict the risk of subsequent events (e.g. injuries) and model their dynamics and relationships to factors such as training,
- Bring together factors including training, competition, injuries, sleep and nutrition, to better understand sensitivities and outcomes in a whole-of-systems model.
- Analyse and predict in-race or in-game dynamics and strategies (e.g. ball and player tracking, pacing and fatigue).
Current partner organisations include:
- Swimming Australia
- Triathlon Australia
- Tennis Australia
- Paddle Australia
- West Coast Eagles
- QUT Sports.
Topics will be catered for the scope of the project (VRES to PhD), but will likely include:
- developed algorithm(s) and/or model(s) and visualisations
- reports and presentations to the centre and industry
- regular engagement (and potentially co-location) with the sports organisation(s).
Skills and experience
- Moderate to strong programming experience, especially with R or Python, is ideal.
- Some prior statistical or machine learning experience is ideal.
- A desire to solve real-world problems is a must.
You may be eligible to apply for a research scholarship.
Contact the supervisor for more information.