AMF: Aggregated Mondrian Forests for Online Learning
This provides an online, parameter-free algorithm for machine learning practitioners needing real-time or streaming data adaptation, though it is incremental as it builds on existing Mondrian Forest methods.
The paper tackles the problem of offline random forests requiring the entire dataset at once by introducing AMF, an online random forest algorithm based on Mondrian Forests, which uses a variant of the Context Tree Weighting to efficiently aggregate over all prunings. Numerical experiments show that AMF is competitive with several strong baselines on many datasets for multi-class classification.
Random Forests (RF) is one of the algorithms of choice in many supervised learning applications, be it classification or regression. The appeal of such tree-ensemble methods comes from a combination of several characteristics: a remarkable accuracy in a variety of tasks, a small number of parameters to tune, robustness with respect to features scaling, a reasonable computational cost for training and prediction, and their suitability in high-dimensional settings. The most commonly used RF variants however are "offline" algorithms, which require the availability of the whole dataset at once. In this paper, we introduce AMF, an online random forest algorithm based on Mondrian Forests. Using a variant of the Context Tree Weighting algorithm, we show that it is possible to efficiently perform an exact aggregation over all prunings of the trees; in particular, this enables to obtain a truly online parameter-free algorithm which is competitive with the optimal pruning of the Mondrian tree, and thus adaptive to the unknown regularity of the regression function. Numerical experiments show that AMF is competitive with respect to several strong baselines on a large number of datasets for multi-class classification.