What is it about?
We select a best subset of features using information theory, especially the notion of complementarity where a subset of features is worth more than the the individual features in the subset. We minimize redundancy, maximize complementarity, while keeping the subset of features as small as possible.
Photo by timJ on Unsplash
Why is it important?
Our feature subset selection helps build better classifiers regardless of the method used for classification.
Read the Original
This page is a summary of: An adaptive heuristic for feature selection based on complementarity, Machine Learning, June 2018, Springer Science + Business Media, DOI: 10.1007/s10994-018-5728-y.
You can read the full text:
The following have contributed to this page