Mean Square Error (MSE) is a measure of the average squared difference between the estimated values and the actual value, serving as a common metric for evaluating the accuracy of a predictive model. It is particularly sensitive to outliers, as larger errors are disproportionately penalized due to the squaring operation.