Add a new reactivity measure: Euclidean distance measured between a scaled
version of the performance curve of ideal conditions (i.e. if the variance for
the current t in the current experiment is 10% LOWER than than the variance for
the current t in ideal conditions, then you would expect to see a 10% INCREASE
in performance for the curve of the current experiment for that point).
Would also be a building block to getting the adaptability measure
implemented. I think.
jharwell
commented
4 years ago
Add a new reactivity measure: Euclidean distance measured between a scaled version of the performance curve of ideal conditions (i.e. if the variance for the current t in the current experiment is 10% LOWER than than the variance for the current t in ideal conditions, then you would expect to see a 10% INCREASE in performance for the curve of the current experiment for that point).
Would also be a building block to getting the adaptability measure implemented. I think.