To learn more about EpiX Analytics' work, please visit our modeling applications, white papers, and training schedule.

Page tree

 

No CrystalBall function

JohnsonU(a1, a2, b, g)

JohnsonU equations

 

 

The main use of the Johnson unbounded distribution is that it can be made to have any combination of skewness and kurtosis. Thus, it provides a flexible distribution to fit to data by matching these moments. That said, it is an infrequently used distribution in risk analysis.

The distribution name comes from Johnson (1949) who proposed a system for categorizing distributions, in much the same spirit that Pearson did. Johnson's idea was to translate distributions to be a function of a unit Normal distribution, one of the few distributions for which there were good tools available at the time to handle.

 


  • No labels