Posterior Probability and Bayes
Jane M. Horgan · 2020
In this chapter, the authors show how Bayes' rule has some important applications in computer science. An important application of Bayesian probabilities is in the area of fault diagnosis in hardware. Bayes' theorem is sometimes used in supervised learning to classify items into two or more groups. Bayes' theorem can be used to estimate the posterior probabilities, that is, the probabilities that an email which is received as spam, really is spam, or is in fact legitimate. Email spam filtering is a way of processing incoming email messages to decide if they are spam or legitimate. Spam detection is an example of the two-case classification problem. The proportions of spam and legitimate emails found in the training set are used as estimates of the prior probabilities. Bayes' rule is sometimes used to help translate text from one natural language to another.