Unverified50% confidenceCautionExact time
The F1 score ignores the number of true negatives, so on extremely imbalanced data it still needs to be judged together with other metrics.
1
Sources
50%
Confidence
Long-term
Relevance
11/7/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedComparing six systems, the paper finds only a weak correlation between verdict accuracy and evidence localization (span F1) scores, with inconsistent ranking of systems across the two metrics.73% similarUnverifiedThe F1 score uses the harmonic mean rather than the arithmetic mean because the harmonic mean is more sensitive to smaller values; as long as either precision or recall is very low, the F1 score is pulled down noticeably.72% similarVerifiedWhen the proportion of diseased samples is low, recall and F1 score reflect the model's actual detection capability better than accuracy.70% similarUnverifiedThe model's F1 score is 72%, and the data has a severe class imbalance with a ratio of 89:11.69% similarUnverifiedClustered standard errors target the situation of non-independent samples and can handle correlation between samples, whereas treating each sample as an independent observation underestimates the variance.69% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1145643API
curl https://kongchang.com/api/v1/knowledge/claims/1145643MCP
get_claim(id=1145643)