Treelib: Decision Trees for Apache SPARK
This repository focus on a possible implementation of decision trees, random forests and a series of "utilities" for Apache SPARK.
- Regression Trees are implemented based on the CART algorithm
- Classification Trees come in two flavors: Binary Classification Trees (implemented based on CART) and Multi-way Classification Trees (implemented based on ID3).
- Pruning follows the methodology ...
- Random forests are implemented using ...
Have a look at our WIKI to learn more about Treelib!