This repo contains a very simple scikit-learn-based implementation of the Random Forest Kernel described in Davies and Ghahramani, 2014. (https://arxiv.org/pdf/1402.4293.pdf) THe main difference is that the authors suggest a stochastic quantification of the tree-induced differences (by selecting a node level from an uniform distribution) while here I query the entire path from the root to the leaves.
What is this repository for?
- Didactic purposes and for future development
How do I get set up?
- Python 2.7 should be installed
- matplotlib, scipy, numpy, scikit-learn should be installed (they can be easily installed with pip)
Who do I talk to?
- My email is in the source file