Machine Learning in Sports 101

AI already helps scoring sporting events, like gymnastics in the upcoming Olympics, thanks to systems like Fujitsu’s for example. Sports science traditionally applied biomechanics for sports analysis, and it works really well! AI can help this traditional approach by adding methods from supervised, unsupervised and reinforcement learning.

Supervised learning can help with labeled sets of data in tasks like classification, for example classifying sports activities:

Unsupervised learning doesn’t assume that data is labeled, instead its goal is finding similarities in the data. It’s often used for self-organizing dimensionality reduction and clustering. For example, if you train an unsupervised model with sufficient data containing images of athletes performing actions in different sports, such a model should be able to predict what group, or sport a given image belongs to.

Reinforcement learning (RL) applies a concept of an agent trying to achieve the goal and receiving a reward for most positive actions. There’re some really cool applications of this method in sports that I covered in my book.

Want to learn more on applying ML in sports? Check out my book and the course that covers a lot of material with detailed walk throughs.