Canonical Correlation Forests (CCFs)

CCFs are a decision tree ensemble method for classification and regression. CCFs naturally accommodate multiple outputs, provide a similar computational complexity to random forests, and inherit their impressive robustness to the choice of input parameters.

This implementation is completely done using Numpy and SciPy, which are open-sourced numerical computing libraries.

CCFs results on Spiral Dataset	CCFs results on Camel Dataset

Prerequisites

1. Numpy == 1.17.3
2. SciPy == 1.3.1
3. Matplotlib == 3.1.2 # For Visualization

(This code base was developed on Python3.6)

Code

pip install -r requirements.txt

For classification example run the following command:

python3 classification_example.py

For regression example run the following command:

python3 regression_example.py

Contribution

Any improvements to the code base are welcomed.

Original Code

https://github.com/twgr/ccfs

Paper citation

@article{rainforth2015canonical,
  title={Canonical correlation forests},
  author={Rainforth, Tom and Wood, Frank},
  journal={arXiv preprint arXiv:1507.05444},
  year={2015}
}