The goal of this project is to develop a framework which is capable of clustering pitches on the basis of their physical characteristics such as speed, movement, release point, and so on. This project is basically composed of three parts as follows:

1. Perform exploratory data analysis including variable selection and unusual value detection
2. Cluster the pitches using k-means clustering
3. Project the pitches onto a two-dimensional space using truncated SVD

1. https://nbviewer.jupyter.org/github/KUANCHENGFU/Pitch-Type-Clustering/blob/main/Pitch_type_clustering.ipynb