InformativeFeatureSelection 3.1.0

Package which provides an feature selection algorithm which considers class separability and an implementation of Informative Normalized Difference Index (INDI)

Install

pip install InformativeFeatureSelection

Features

• Several implementations of feature selection algorithms based on discriminant analysis
• Binary implementation of Informative Normalized Difference Index (INDI)
• Multiclass implementation of INDI

INDI may be extremely usefully in hyperspectral imaging analysis.

Implemented algorithms were proposed in the following papers:

1. Paringer RA, Mukhin AV, Kupriyanov AV. Formation of an informative index for recognizing specified objects in hyperspectral data. Computer Optics 2021; 45(6): 873-878. DOI: 10.18287/2412-6179-CO-930.

2. Mukhin, A., Paringer, R. and Ilyasova, N., 2021, September. Feature selection algorithm with feature space separability estimation using discriminant analysis. In 2021 International Conference on Information Technology and Nanotechnology (ITNT) (pp. 1-4). IEEE.

Requirements

To simplify usage, two Docker images were created:

1. banayaki/feature-selection:base This image serves as a base image for InformativeFeatureSelection. It includes the necessary python and its packages.

2. banayaki/feature-selection:notebook This image serves as an extension of the base image for InformativeFeatureSelection. It includes an additional tool: Jupyter Notebook. The Jupyter server starts automatically when the container begins.

How to use them?

Just run the following command:

docker container run --rm -p 8888:8888 -v ./project:/home/workdir banayaki/feature-selection:notebook

Then just copy jupyter's token from container's log.

Usage example

See jupyter notebook file in examples folder.

License

MIT License