OVERVIEW
Implementation of graph based diffusions. Heat Kernel, Personalised PageRank, and custimized diffusions provided. Useful for semi-supervised learning
CONTAINS
Module gpmod contains:
a) predict.py
Class definition of diffusion_SSL() graph-based diffusions,
with member functions for..
-
Loading a graph adjacency matrix
-
Configuring diffusion parameters
-
Seeding and running diffusion
b) utils.py
Some utility functions
USAGE
See example.py and documentation in predict.py