Expected value
Also: expectation, mean, E[X]
E[X] = Σ x·P(X = x) for discrete X, or ∫ x·f(x) dx for continuous X. Losses, returns in reinforcement learning and many training objectives are expected values.
E[X] = Σₓ x · P(X = x)
Example
For a fair die, E[X] = (1 + 2 + 3 + 4 + 5 + 6) / 6 = 3.5, even though no single roll can show 3.5.