Jensen-Shannon Divergence and Machine Learning Toolbox
The JSDML is a python implementation of a neural network using Keras/Tensorflow which also includes functionality to calculate the JSD between features and classes. Some of the functionality is borrowed from Greg Ver Steeg's package NPEET (https://github.com/gregversteeg/NPEET).
A quick start guide and full documentation are included along with examples. Any questions can be directed to the authors at [email protected].