Semi-Supervised Learning (5.0.4)

Choice of the Learning Algorithm

Semi-Supervised Learning is an Unsupervised learning restricted to the nodes that are within in the Target node radius

Those nodes are found by applying the Markov Blanket algorithm on the Target node, and then, iteratively on the nodes belonging to the Target node’s Markov Blanket.

This allows carrying out unsupervised learning even when there are a lot of variables, e.g. for Microarray analysis.

The structural learning algorithm applied on the automatically selected nodes can now be selected among the 3 naïve based algorithms and the 5 unsupervised learning algorithms