Bayes' theorem provides the formal mathematical machinery for inverting conditional probabilities.
It calculates the probability of an unobserved cause given an observed effect.
The theorem expresses posterior probability as prior probability multiplied by likelihood, divided by evidence.
The prior probability represents the initial belief in a hypothesis before collecting current evidence.
The likelihood quantifies the probability that the observed evidence would appear if the hypothesis were indeed true.
The evidence, or marginal probability, acts as a normalising constant ensuring that posterior probabilities sum to one.
It calculates the total probability of observing the evidence by summing likelihoods across all competing hypotheses.
For example, in automated spam detection, the theorem calculates the posterior probability that an email is spam given observed words.
If a specific phrase appears frequently in spam but rarely in legitimate messages, its presence boosts the posterior spam probability.
In clinical medicine, Bayes' theorem reveals why screening tests for rare diseases can produce high rates of false positives.
When a disease has low prior prevalence, even an accurate diagnostic test yields many false alarms across the healthy majority.
Bayesian inference treats parameters as random variables with probability distributions rather than fixed unknown constants.
As new data points arrive sequentially, today posterior distribution serves as tomorrow prior distribution.
A key practical challenge in Bayesian analysis is selecting an appropriate prior distribution that reflects genuine baseline knowledge.
A frequent mistake is confusing the likelihood of evidence given a hypothesis with the posterior probability of the hypothesis itself.