chenxofhit/fast-mRMR
An improved implementation of the classical feature selection method: minimum Redundancy and Maximum Relevance (mRMR).
Full Featured HTML Framework For Building iOS & Android Apps
Public repository record indexed from GitHub. Explore verified star velocity metrics, source code repositories, and curated developer tool directories across the GitHubRepo ecosystem.
An improved implementation of the classical feature selection method: minimum Redundancy and Maximum Relevance (mRMR).
libLBFGS: a library of Limited-memory Broyden-Fletcher-Goldfarb-Shanno (L-BFGS)
Mirror of Apache Spark
A MATLAB interface for L-BFGS-B