Yet Another Random Forest ========================= An implementation of the random forests machine learning algorithm in C++. Currently supports classification only, not fully tested. Features include: Class probability distributions at each node (as opposed to just single votes). OOB error rates. Feature importance using a permutation test on the OOB samples. In progress: Image segmentation/classification, currently some Haar-like features are available. Aims: Configurable split rules (currently only one: single feature threshold scored using information gain).