Learning - Multiple Clustering (5.0.2)

It is possible to use the BayesiaLab Semi-Supervised Clustering algorithm for Multiple Clustering. The weights of the variables are defined with the Relative Significance of their relation with the Target node, either directly, when the Weighting coefficient is equal to 1, or partially otherwise.

This option can help having Latent variables more correlated to the Target Node. Indeed, the Factor's variables that are weakly correlated with the Target node will slightly participate to the clustering.