Recall@K
Recall@K: The share of relevant items retrieved within the top k results relative to all available relevant items. It is common in retrieval and recommendation. It measures neither exact ranking position nor quality of irrelevant results shown. This perspective connects the definition with practical behaviour in artificial intelligence systems.
In italiano: Recall@KLegge V · Overfitting, bias e collasso dei modelliIt is common in retrieval and recommendation. It measures neither exact ranking position nor quality of irrelevant results shown.