Tantárgy adatlapja
Description of the subject:
This advanced-level course focuses on the processing, analysis, and modeling of sport and rehabilitation data using state-of-the-art data science methodologies. The course blends theoretical foundations with practical applications, emphasizing real-world data sets from wearables, motion capture systems, and clinical rehabilitation monitoring tools. Students will apply statistical models, machine learning, and deep learning to extract insights from time-series and biomechanical data. Emphasis is placed on the design and implementation of personalized sport-performance tracking systems, injury prediction, recovery modeling, and performance forecasting.
Planned weekly schedule/topics (indicative):
- Introduction to Sport Data Ecosystems and Ethics
- Advanced Sensor Technologies in Sports & Rehabilitation
- Signal Processing for Biometric and Kinematic Data
- Time-Series Analysis and Dynamic Modeling in Sports
- Functional Human Anatomy Revisited (with application to modeling)
- Advanced Statistical Methods for Sports Data
- Feature Engineering from Wearable and Movement Data
- Machine Learning for Performance Optimization and Injury Risk Assessment
- Deep Learning Architectures in Sport Applications (CNNs, RNNs, Transformers)
- Computer Vision in Sports: Pose Estimation and Action Recognition
- Complex Systems and Network Dynamics in Team Sports
- Multimodal Data Integration (EMG, ECG, GPS, Video)
- Building End-to-End Sport Analytics Pipelines
- Sport Forecasting Models and Decision Support Systems
- Final Project Presentations and Case Studies
List of required and recommended literature:
- Thomas W. Miller – Sports Analytics and Data Science: Winning the Game with Methods and Models, Pearson, ISBN: 9780134193886
- Mark Powell – Sport Technology: History, Philosophy and Policy, Routledge, 2017, ISBN: 9781138844321
- László Grand et al. – Deep Learning for Movement Analysis (lecture notes, downloadable)
- Wilkerson et al. – “Machine Learning for Injury Risk Prediction in Athletes”, Journal of Sports Sciences, 2020
- Open-access datasets and publications from the MIT Human Dynamics Lab and FIFA Innovation Programme
List of required and recommended skills:
- Application of machine learning and deep learning in real-world sport scenarios
- High-level data processing and signal interpretation from multiple biosensor types
- Ability to analyze and model complex dynamic systems (e.g., team play, recovery models)
- Integration of anatomical knowledge into predictive and diagnostic modeling
- Development of end-to-end analytic pipelines for applied sport and rehab contexts
- Scientific thinking, interpretation of literature and data for practical innovation
