MotherTree: Meta-learning on synthetic data improves decision tree training
Conventional decision tree algorithms produce effective, transparent models that can be audited, communicated, and deployed independently of the training data, but require learning every new task from scratch.
arXiv cs.LG2 min read