no code implementations • 3 Mar 2023 • Shir Friedman, Gonen Singer, Neta Rabin
We aim to drastically reduce the number of selected features, in order to create a sketch of the original data that codes valuable information for the classification task.