Abstract
The problem of selecting features for a data set with a small number of objects is one of the most complex ones. Significant features selected for such data sets can vary quite a lot depending on how sub-sampling was performed during validation. This effect is called low feature set stability and signals on low reliability of the selected features. We propose a feature selection algorithm that is based on bagging procedure of feature selection filters quality measures ensemble and allows to obtain more stable feature sets, than would be obtained by running conventional algorithms, called BagMeLiF. This algorithm is based on MeLiF algorithm and will outperform original algorithm both in F1 score and stability with hyperparameter k around 0.7–0.9 if the dataset is well-balanced, but if it is not, then k around 0.1–0.2 will the best which is a quite straightforwardly applicable result.
Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
