aka risk, cost, loss; or its opposite, utility, reward,... depending on the area or context
A loss function refers to a Function that one wishes to minimize, in some decision theoretic framework, like in learning, Operations research, mathematical Ethics, etc., Optimization. Its oppositve (negative) is the Utility function, which one wishes to maximize.
Absolute value loss
Epsilon insensitive loss function
Training error The problem is that this is not convex. That is why Hinge loss is better.
Negative log likelihood
See Empirical risk minimization
Expected loss, defined over Measurable functions
| I | Ii |
| II | L |