Swift ML support is on the horizon (checkout Swift TensorFlow, swift-jupyter). Although swift is still in it's primitive stage, the possibility of utilizing a fast, modern programming language in machine learning is fascinating.
This project will help me re-learn Swift by implementing this well-known statistical model which is also the building block for the versatile Random Forest learning algorithm.
In this project, I want to achieve several goals:
- Implement Decision Tree learning using CART.
- Evaluate performance of the implementation on several datasets.
- Demonstrate how to use JupyterLab along with Swift kernel by leveraging the swift-jupyter project. (This cannot be done easily right now, as the swift-jupyter notebook is still not quite stable, and the "%include" directive doesn't work as expected).