.BusinessSimulation.MoleculesOfStructure.InformationProcessing.AggregatePerformance

Information

This information is part of the Business Simulation Library (BSL). Please support this work and ► donate.

The output y is given by applying an aggregate function (e.g., →Min, →GeometricMean, →ArithmeticMean, →Max) to a set of performance indicator inputs u. Optionally a vector of weights can be given as constant weights or external input u_weights, so that weighted averages can be used.

Notes

• Since ultimately very different aspects of a system's performance need to be monitored, aggregated, and compared, it is advisable to use dimensionless performance scores—if possible in the closed unit interval [0,1] which generalizes the space of utility functions.
  
  
• It it well established, that the following ordering holds in general: max(u) ≤ geometricMean(u) ≤ arithmeticMean(u) ≤ max(u)
  
  
• If deficits in any of the inputs should not be easily compensatable by surplusses in other inputs, then the geometric mean should be preferred to the arithmetic mean (the minimum would obviously be the extreme case of any deficit dominating the calculation of an aggregate).

See also

BasicOrientation, PerformanceIndicator