normalizedMI: Computes the normalized mutual information

In RobustEM: Robust Mixture Modeling Fitted via Spatial-EM Algorithm for Model-Based Clustering and Outlier Detection

Description

This uses the normalized mutual information to evaluate how well the EM algorithm performs with respect to the data provided.

Usage

normalizedMI(trueLabel, predictedLabel)

Arguments

trueLabel	This is a vector that represents the actual cluster labels of the data
predictedLabel	This is the predicted cluster labels returned by the EM algorithm

Value

normalizedMI

This value represents the efficiency of the algorithm. The higher the value, the greater the efficiency of the algorithm

Warning

This function is dependent on the true and predicted clusters having the same levels.

References

Manning, C., Raghavan, P., Schutze, H. (2008). An Introduction to Information Retrieval. Cambridge University Press. ISBN 0-521-86571-9

See Also

matchCluster

Examples

## Not run: 
  true<-c("a","b","a","a","a")
   predicted<-c("a","b","b","a","a")
   normalizedMI(true,predicted)

## End(Not run)

RobustEM documentation built on April 14, 2017, 10:05 a.m.