NeurIPS 2020

Sparse Learning with CART

Meta Review

The reviewers generally agreed that this was a well-written and comprehensive paper that makes a valuable contribution. While there was some uncertainty about how significantly this work will motivate future research, this was not considered to be a strongly negative point. The main concerns were the lack of experiments (R2 and R4), and the limited discussion of ensembles (R1). The author response did a great job in addressing these concerns: all three of these reviewers seem to be mostly satisfied with the promised changes. Overall, this is a solid paper, and the changes proposed in the author response should make it even better. In terms of smaller issues, please carefully read the reviews, and take their suggestions seriously when making edits.