Compute overall accuracy and per-class accuracy from a confusion matrix, for model evaluation and error analysis reports.
类别不均衡时准确率易误导,应结合其他指标。
| Item | Formula / Note | Source / Standard |
|---|---|---|
| Accuracy formula | accuracy is the trace of the confusion matrix divided by the total sample count | ISO/IEC 22989:2022 AI concepts |
| Per-class accuracy | per-class accuracy is the diagonal cell over the row total which is class recall | ISO/IEC TR 24028:2020 AI trustworthiness |
| Class imbalance | under class imbalance overall accuracy is dominated by the majority class | Sokolova & Lapalme 2009, Information Processing |