Forgetful Forests: Data Structures for Machine Learning on Streaming Data under Concept Drift

Database and data structure research can improve machine learning performance in many ways. One way is to design better algorithms on structures. This paper combines the use of incremental computation as well sequential probabilistic filtering enable “forgetful” tree-based cope with streaming that suffers from concept drift. (Concept drift occurs when functional mapping input classification changes over time). The forgetful described this achieve high while maintaining quality predictions data. Specifically, are up 24 times faster than state-of-the-art with, at most, a 2% loss accuracy, or least twice without any accuracy. makes such structures suitable for volume applications.