About
The feature selection repository is designed to collect some widely used feature selection algorithms that have been developed in the feature selection research to serve as a platform for facilitating their application, comparison and joint study. The feature selection repository also effectively assists researchers to achieve more reliable evaluation in the process of developing new feature selection algorithms. We develop the open source feature selection repository scikit-feast by one of the most popular programming language - python. It contains more than 40 popular feature selection algorithms, including most traditional feature selection algorithms and some structural and streaming feature selection algorithms. It is built upon one widely used machine learning package scikit-learn and two scientific computing packages Numpy and Scipy.
Installing scikit-feast
Prerequisites:
Steps:
After you download scikit-feast-1.0.0.zip, unzip the archive. Then the scikit-feast root directory will contain a setup script setup.py and a file named README.txt.
For Linux, under the scikit-feast root directory building and installing the module distribution is a simple matter of running one command from a terminal:
python setup.py install
For Windows, this command should be run from a command prompt window (Start->Accessories), under the scikit-feast root directory:
setup.py install
Source code is available on scikit-learn GitHub repository.
For scikit-feast API usage, please refer scikit-learn feature selection repository API Document.
A short introduction into scikit-learn feature selection repository scikit-learn feature selection tutorial
Citing
If you find scikit-feast feature selection reposoitory useful in your research, please consider citing the following paper [pdf] :
@article{Li-etal16,
title= {Feature Selection: A Data Perspective},
author= {J. Li and K. Cheng and S. Wang and F. Morstatter and R. Trevino and J. Tang and H. Liu},
organization= {Arizona State University},
year= {2016},
url= {http://featureselection.asu.edu/scikit-feast},
}
Contact
Jundong Li
E-mail: jundong.li@asu.edu
Kewei Cheng
E-mail: kcheng18@asu.edu
Huan Liu
E-mail: huan.liu@asu.edu
Physical address: Brickyard Suite 553 (CIDSE), 699 South Mill Ave, Tempe, AZ 85281