I’m working on a project that involves analyzing player movement in field sports (like soccer or hockey), and I’m looking for machine learning solutions to help with this. Specifically, I’m interested in methods or tools that can track and analyze player positions, movement patterns, and possibly predict future movements based on historical data.
Has anyone worked on similar projects or have recommendations for machine learning techniques or frameworks that could be effective for this type of analysis? Any insights on data collection methods, feature engineering, or model selection would also be greatly appreciated.
For hockey, we used a mix of LSTM networks and convolutional layers to predict player movements based on past data. Having a reliable dataset was crucial, so we spent time developing a solid system to label player positions in the game footage accurately.
