Evaluation & Quality
Aleatoric uncertainty
Uncertainty caused by natural variation or unclear information, rather than simply not having learned enough.
Example
Two similar situations produce different outcomes even when measured in the same way.
Why people use it
It reminds people that some variation cannot be removed simply by building a smarter predictor.
What you'll hear
“Some variation remains even when the information is good.”
What this means for you
Do not expect an AI system to predict inherently variable outcomes perfectly.
Can you control it?
No
No direct control. This describes a wider issue, concept or result rather than something you can simply switch on or off in a tool.
Common questions
- Can this uncertainty differ between cases?
- Yes. Some situations naturally vary more than others, even when the AI has similarly useful information about each.
- Will more examples remove all of it?
- No. More examples can improve understanding without eliminating variation that remains in the situation.
- Can better measurements sometimes reduce it?
- Yes. What appears unavoidable may partly reflect details the current measurements miss.