Cluster Interpretation: Posterior Distributions

Background & Context

• On this page, we present Posterior Distributions for cluster interpretation as an alternative to Most Relevant Explanations for Cluster Interpretation.

• To provide further context for Most Relevant Explanations for Cluster Interpretation, we compare several other approaches that can help interpret individual Clusters:

  • Setting Evidence for Cluster Interpretation: Posterior Distributions, Relationship with Target Node, Mosaic Analysis, Posterior Mean Analysis, Segment Profile Analysis, Histograms, Tornado Diagrams,
  • Optimization for Cluster Interpretation: Dynamic Profile, Target Optimization Tree

• More specifically, we compare all these approaches with regard to characterizing the state $\mathit{Cluster\ 3}$ of the Cluster Node $\mathit{Factor}_0$ in the reference network.

• All analyses and instructions on this page refer to this reference network, which you can download here:

  MaleClusters.xbl

Posterior Distributions for Cluster Interpretation

• We bring up all of the network’s nodes by selecting Monitor Panel Context Menu > Sort > Target State Correlation.
• In addition to displaying all Monitors, they are also automatically ordered in terms of their importance with the Target State $\mathit{C3}$.

• The obvious starting point of our exploration is to set the evidence on the factor node that corresponds to our interest in $\mathit{C3}$, i.e., $\mathit{Factor}_0=\mathit{C3}$.

• In addition to showing the posterior distributions of the nodes, BayesiaLab can display Absolute Variations: Monitor Context Menu > Absolute Variations. This highlights the difference in the distributions of $\mathit{C3}$ members versus the marginal distributions.