Aleatoric Uncertainty
The inherent, unavoidable and irreducible randomness in a dataset that cannot be eliminated simply by collecting more data or building a more complex model.
Think of It Like This
Like flipping a fair coin; no matter how closely you study the coin's physics or how many times you flip it, you still can't perfectly predict the next outcome.
Unlike epistemic uncertainty, which stems from the model's ignorance and shrinks with better training, aleatoric uncertainty is baked into the problem itself. It arises from noisy sensors, hidden variables, or genuinely stochastic processes. Knowing this distinction stops teams from wasting immense compute trying to 'solve' noise that fundamentally cannot be predicted.