Feature selection with neural estimation of mutual information
Primary Area: general machine learning (i.e., none of the above)
Code Of Ethics: I acknowledge that I and all co-authors of this work have read and commit to adhering to the ICLR Code of Ethics.
Keywords: Feature selection, mutual information, MINE, neural networks
Submission Guidelines: I certify that this submission complies with the submission instructions as described on https://iclr.cc/Conferences/2024/AuthorGuide.
TL;DR: Novel feature selection method based on neural estimation of mutual information
Abstract: We describe a novel approach to supervised feature selection based on neural estimation of mutual information between features and targets. Our feature selection filter evaluates subsets of features as an ensemble, instead of considering one feature at a time as most feature selection filters do. This allows us to capture sophisticated relationships between features and targets, and to take such sophisticated
relationships into account when selecting relevant features. We give examples of such relationships, and we demonstrate that in this way we are capable of performing an exact selection, whereas other existing methods fail to do so.
Anonymous Url: I certify that there is no URL (e.g., github page) that could be used to find authors' identity.
No Acknowledgement Section: I certify that there is no acknowledgement section in this submission for double blind review.
Submission Number: 112
Loading