Child pages
  • Most Probable Explanation

Contents

This tool allow you to compute the Most Probable Explanation, i.e. the case that has the highest joint probability in a given network. The Monitors highlight this case. The probability of each state is replaced by the likelihood that the corresponding state belongs to most probable case. The context (i.e. the set evidence) is taken into account during the analysis. Each time a piece of evidence is set, the Most Probable Explanation is recomputed and the Monitors are updated accordingly. The joint probability displayed in the upper part of the window of the Monitors corresponds to the joint probability of this Most Probable Explanation


 

The means and the standard deviations of the continuous nodes are no longer computed and displayed. The same applies to the Mean Value, the Total Value and, if necessary, the uncertainty and the likelihood, which are no longer computed and displayed in the information panel.