Multimodal AI
Mean average precision
A combined score measuring how well systems find relevant items across groups or tasks.
Example
A picture detector's scores are combined across several object groups.
Why people use it
It combines detailed precision scores into a summary for tasks such as finding objects or search results.
What you'll hear
“What does this average include?”
What this means for you
Read the matching and averaging rules before comparing reported scores.
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 scores from different tests be compared directly?
- Only cautiously. Different groups, matching rules and averaging methods can make the numbers mean different things.
- Can the average hide a weak category?
- Yes. Strong performance elsewhere can conceal poor results for a particular kind of item.
- Does one extra correct detection always change the score equally?
- No. Its rank and the evaluation rules affect its contribution.