This image shows the formula for Mean Square Error, also known as MSE. For each value from i to n, the difference beween the actual likelihood and the corresponding predicted likelihood is squared, then divided by n. The n results are then accumulated, which gives the Mean Square Error.