Semi-Supervised Learning
- Semi-Supervised Learning is an unsupervised learning algorithm that searches the relationships between the nodes that are within a predefined distance of the target.
- This distance is computed by using the Markov Blanket learning algorithm.
- The semi-supervised learning algorithm allows learning a network fragment centered on the Target Node.
- This algorithm is very useful for learning networks with a large number of nodes, e.g., in the context of microarrays analysis with thousand of genes.
- Furthermore, Semi-Supervised Learning is practical for generating predictive models as an alternative to learning Markov Blanket in case the nodes in the Markov Blanket contain missing values. Nodes with missing values would not separate the Target Node from other nodes in the network.
In addition to the default learning options, you can specify the search depth from the Target Node: