BAYES' THEOREM CaDDiS Home Page



The fundamental result that expresses a conditional probability P(A|B) of an event A given an event B as

 

More generally, where  is one of a set of I events which together make up all possible eventualities (such as all of the 20 diseases under consideration),

 

This allows estimates of probability to be continually revised in the light of new observations. (Named after the English probabilist and theologian Thomas Bayes (1702-61)).
 
 

e.g. We may have an estimate of our belief in whether a cow is infected with ECF given that we have observed that it has enlarged lymph nodes: P(ECF | Lymph nodes enlarged). If we now observe that the animal is anaemic, this will change our belief in whether it has the disease. Our new belief can be calculated using Bayes' theorem.