This policy aggregates a set of discrete-valued responses into a single best response using the adversarial scheme outlined in the AutoMan OOPSLA '12 paper.
Find the multiple-comparisons-adjusted confidence level using the Bonferroni-Holm adjustment.
Find the multiple-comparisons-adjusted confidence level using the Bonferroni-Holm adjustment.
The unadjusted confidence level.
The number of comparisons.
The adjusted confidence level.