BFS documentation

BFS is a Python library for feature selection based on a constrained biobjective gradient descent algorithm. It is based on network sparsification using a sparse group lasso regularizer. It is tailored for High Dimension Low Sample Size data. In addition to the classification results, BFS also returns the list of feature in order of importance.

Check out the Installation and Usage sections for further information, including how to install the project.

Note

This project is under active development.

Contents

• Installation
• Usage
• BFS
  • Arguments
  • Methods
• Utils
  • Methods