What is a 'Loss Function'?

What is a 'Loss Function'?

Approach: Define the mathematical measure of a single prediction error.

A function that maps an event or values of one or more variables onto a real number intuitively representing some 'cost' associated with the event. In ML, it calculates the error for a single training example.

Follow-up: What is the difference between loss and cost function?

Key concepts: error, objective function, gradients.