This is a simple random forest classifier implementation in Java following the notes from the technical report from MSR. Included is an axis-aligned learner which is tested on two synthetic datasets: a four-class uniformly distributed dataset and a five-class spiral dataset. The unit tests for the axis-aligned learner also generate some nifty visualizations (roughly plotted via Java 2D), which are included below:
This implementation was originally created to help me understand how random forests work. It's not optimized for performance.
Project setup uses gradle. To get started, checkout and run,
./gradlew test