ABSTRACT
This chapter reviews Bayes Theorem and illustrates different applications of it through examples from chromatography. Bayes Theorem unites two basic rules of probability, the product rule and the addition rule, and tells us how to treat conditional probabilities. The use of Bayes Theorem to estimate parameter values is possible because the set of parameter values that maximize their posterior probability are, ipso facto, the best estimates. The most used Bayes classifier is the "naive" Bayes classifier, which covers different approaches, some of which are not strictly Baye, but all assume independence between features. An advantage of naive Bayes is that it can estimate the parameters necessary for classification from minimum training data. The performance of a field test kit for cholera using a dipstick immunochromatographic test, also known as a rapid diagnostic test (RDT), was evaluated without a "gold standard" by Bayesian latent class modeling.
