Loss function

In statistical decision theory, a function measuring the cost of estimating a parameter with a value that deviates from the truth. Common examples include squared error, absolute error, and 0-1 loss. Bayes estimators are derived by minimizing the expected loss, known as the Bayes risk.