Estimates the probability of a mismatch under a Bayesian model with Binomial likelihood and Beta(a, b) prior.
Estimates the probability of a mismatch under a Bayesian model with Binomial likelihood and Beta(a, b) prior. When a = b = 1, this is also known as "Laplace's rule of succession".
TODO: Beta(1, 1) is the safest choice, but maybe Beta(1/2, 1/2) is more accurate?
Empirically estimated probability of a mismatch.
Empirically estimated probability of a mismatch, as a QualityScore.
An empirical frequency count of mismatches from the reference.
This is used in ObservationTable, which maps from CovariateKey to Observation.