Abstract
Techniques are described herein to efficiently detect redundant features in a machine learning process. The techniques are able to compute feature redundancy not only for a single feature at a time, but for any subset of features without the need to naively train and evaluate a classifier for each combination of features.
Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.
Recommended Citation
Komarek, Tomas; Brabec, Jan; and Machlica, Lukas, "MACHINE LEARNING FOR THING DETECTION, BEHAVIOR ANALYTICS, AND THREAT DETECTION: AN EFFICIENT ELIMINATION METHOD OF REDUNDANT FEATURE-SUBSETS", Technical Disclosure Commons, (October 01, 2018)
https://www.tdcommons.org/dpubs_series/1540