Skip to Content
BayesiaLabBayesiaLab User GuideMenusAnalysisReportInformation

Information Analysis

Overview

Clicking the Select Nodes buttons defines three sets of nodes (X, Y, and Z) for analysis. The selection is done by directly selecting the nodes in the graph. By default, the Target node is associated with the Y set.

2981986 2981987

Information Analysis Report

Given the exponential growth of the state space, this analysis is performed only on the database.

2981990

Unconditional Entropy

H(X)=−∑x∈Xp(x)log⁡2(p(x))\displaystyle H(X) = -\sum_{x \in X} p(x) \log_2(p(x))

2981992 2981993

Conditional Entropy

H(X∣Z)=−∑z∈Zp(z)∑x∈Xp(x∣z)log⁡2(p(x∣z))\displaystyle H(X \mid Z) = -\sum_{z \in Z} p(z) \sum_{x \in X} p(x \mid z) \log_2(p(x \mid z))

2981994 2981996

Unconditional Mutual Information

I(X,Y)=H(X)−H(X∣Y)\displaystyle I(X, Y) = H(X) - H(X \mid Y)

2981997

Conditional Mutual Information

I(X,Y∣Z)=H(X∣Z)−H(X∣Y,Z)\displaystyle I(X, Y \mid Z) = H(X \mid Z) - H(X \mid Y,Z)

2981999 2982001 2982002

Total Correlation

C(X)=∑iH(xi)−H(X)\displaystyle C(X) = \sum_i H(x_i) - H(X)

2981999

Conditional Total Correlation

C(X∣Z)=∑iH(xi∣Z)−H(X∣Z)\displaystyle C(X \mid Z) = \sum_i H(x_i \mid Z) - H(X \mid Z)

2981999
Example: Most Relevant Explanations for Failure AnalysisHidden Variable Discovery