Evaluation & Quality
Accuracy
The share of predictions or answers that are correct.
Example
A classifier gets 920 of 1,000 labeled examples correct, giving it 92% accuracy.
Why it matters
It gives a simple starting measure of how often a system gets a decision right.
You'll hear it
“The overall score looks good, but it misses the rare cases.”
Can you control it?
Set by the team
The people building or running the AI choose this setup. An everyday user generally needs their help to change how this part works.
Common questions
- Can high accuracy hide a useless detector?
- Yes. Always choosing a very common group can look accurate while missing the cases that matter.
- Does it show how costly mistakes are?
- No. A wrong answer counts as wrong regardless of its practical consequences.
- Can accuracy figures from different tests be compared directly?
- Only cautiously. Different examples and scoring rules can make the percentages mean different things.