This toolkit consists of implementations of various graph-based semi-supervised learning (SSL) algorithms. Currently, three algorithms are implemented: Gaussian Random Fields (GRF), Adsorption, and Modified Adsorption (MAD). Junto also contains Hadoop-based implementations of these three algorithms. https://github.com/parthatalukdar/junto

Label propagation is a popular approach to semi-supervised learning in which nodes in a graph represent features or instances to be classified and the labels for a classification task are pushed around the graph from nodes that have initial label assignments to their neighbors and beyond.

Information

• GitHub: https://github.com/scalanlp/junto