@ Riikka Huusari, 2021

https://arxiv.org/abs/2108.12199

This directory contains implementation of the proposed Sparse Pre-image Kernel Machine (SPKM):
* spkm.py contains the algorithm
* kernels_and_gradients.py contains kernel implementations compatible with spkm code
* losses_and_gradients.py contains loss function implementations compatible with spkm code
* classification_illustration.py and regression_illustration.py provide the illustrations in the paper and show examples of how to use the algorithm