ROC曲線・AUC解析
AUC・感度・特異度・閾値表
ROC曲線・AUC解析
AUC・感度・特異度・閾値表
計算はこのブラウザ内で実行されます列の順序: actual_1_or_0, prediction_score
編集可能な表
| actual_1_or_0 | prediction_score | |
|---|---|---|
CSV / TSV を貼り付け
解析結果
| Threshold | Sensitivity | Specificity | TP | FP | TN | FN | Youden J |
|---|---|---|---|---|---|---|---|
| 0.92 | 0.25 | 1 | 1 | 0 | 4 | 3 | 0.25 |
| 0.81 | 0.5 | 1 | 2 | 0 | 4 | 2 | 0.5 |
| 0.74 | 0.75 | 1 | 3 | 0 | 4 | 1 | 0.75 |
| 0.66 | 0.75 | 0.75 | 3 | 1 | 3 | 1 | 0.5 |
| 0.61 | 1 | 0.75 | 4 | 1 | 3 | 0 | 0.75 |
| 0.42 | 1 | 0.5 | 4 | 2 | 2 | 0 | 0.5 |
| 0.31 | 1 | 0.25 | 4 | 3 | 1 | 0 | 0.25 |
| 0.12 | 1 | 0 | 4 | 4 | 0 | 0 | 0 |
⚠ The displayed threshold maximizes Youden J in this dataset; choose clinical or experimental cutoffs using a prespecified validation strategy.
Data requirements
Enter actual class (1/0) and numeric prediction score.
actual_1_or_0, prediction_scoreReview before reporting
研究用途のみ。原データと解析計画で検証してください。
Check raw observations, excluded rows, model assumptions and study-specific acceptance criteria before using the result.
Transparent analysis
Inputs are processed locally in your browser. Review the calculation policy, limitations and verification approach before reporting a result.
Read the methodology →