all AI news
Rolling Lookahead Learning for Optimal Classification Trees. (arXiv:2304.10830v1 [cs.LG])
cs.LG updates on arXiv.org arxiv.org
Classification trees continue to be widely adopted in machine learning
applications due to their inherently interpretable nature and scalability. We
propose a rolling subtree lookahead algorithm that combines the relative
scalability of the myopic approaches with the foresight of the optimal
approaches in constructing trees. The limited foresight embedded in our
algorithm mitigates the learning pathology observed in optimal approaches. At
the heart of our algorithm lies a novel two-depth optimal binary classification
tree formulation flexible to handle any loss …
algorithm applications arxiv binary classification embedded function lies loss machine machine learning machine learning applications nature novel scalability tree trees