Additional file 1 of Comparison of dynamic updating strategies for clinical prediction models
Notice bibliographique
Résumé
Additional file 1: Table 1. Sample size, number of deaths, and mortality risk per quarter in the post-baseline period. Figure 1. Proportion of deaths versus time in the post-baseline period. Figure 2. Overall calibration for each update strategy when update interval equals 1 quarter and 100% new data are used. Triangles represent deciles of predicted probabilities; vertical lines at bottom of plot represent the distribution of subjects stratified by outcome status (below x-axis = alive, above x-axis = dead). Figure 3. Brier Score versus time when update interval equals 1 quarter and 100% new data are used. Figure 4. AUC vs. time when update interval equals 1 quarter and 100% new data are used. Figure 5. Hosmer-Lemeshow statistic vs. time when update interval equals 1 quarter and 100% new data are used. Figure 6. Calibration vs. time when update interval equals 1 quarter and 100% new data are used. A) calibration intercept (calibration-in-the-large); B) calibration slope. Table 2. Pairwise comparisons (Wilcoxon Signed Rank Test) of Brier Score (BS) at each quarter when the update interval equals 1, 2, 4, or 8 quarters and 100% new data are used for each update. Values represent the median difference (p-value) in BS between the strategy listed in the row header and that listed in the column header. Negative differences indicate that the row header has lower BS (performs better) than the column header; positive differences indicate that the row header has higher BS (performs worse) than the column header. Table 3. Pairwise comparisons (Wilcoxon Signed Rank Test) of Brier Score (BS) at each quarter when the sliding window equals 1, 2, 4, or 8 quarters and the update interval equals 1 quarter. Values represent the median difference (p-value) in BS between the strategy listed in the row header and that listed in the column header. Negative differences indicate that the row header has lower BS (performs better) than the column header; positive differences indicate that the row header has higher BS (performs worse) than the column header. Table 4. Brier Score (BS), AUC, Hosmer-Lemeshow statistic, calibration intercept and slope (from logistic calibration), and number of updates performed under each update strategy when the sliding window equals 4 or 8 quarters and the update interval equals 4 quarters. Table 5. Number of times each candidate update model was selected by the Closed Testing Procedure when the sliding window equals 4 or 8 quarters and the update interval equals 4 quarters. Table 6. Pairwise comparisons (Wilcoxon Signed Rank Test) of Brier Score (BS) at each quarter when the sliding window equals 4 or 8 quarters and the update interval equals 4 quarters. Values represent the median difference (p-value) in BS between the strategy listed in the row header and that listed in the column header. Negative differences indicate that the row header has lower BS (performs better) than the column header; positive differences indicate that the row header has higher BS (performs worse) than the column header. Figure 7. Boxplot of difference in Brier Score between each update strategy and never update calculated at each quarter versus sliding window length, when the update interval equals 4 quarters. Horizontal bars represent medians and interquartile ranges (IQR). Negative values reflect improvement compared never update. Figure 8. Overall calibration for each update strategy when update interval equals 4 quarters and 100% new data are used. Triangles represent deciles of predicted probabilities; vertical lines at bottom of plot represent the distribution of subjects stratified by outcome status (below x-axis = alive, above x-axis = dead). Figure 9. Brier Score versus time when update interval equals 4 quarters and 100% new data are used. Figure 10. AUC vs. time when update interval equals 4 quarters and 100% new data are used. Figure 11. Hosmer-Lemeshow statistic vs. time when update interval equals 4 quarters and 100% new data are used. Figure 12. Calibration vs. time when update interval equals 4 quarters and 100% new data are used. A) calibration intercept (calibration-in-the-large); B) calibration slope.
Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.
Comment cette classification a été obtenuedéplier
Prédiction machine sur la base complète
Imitation des enseignantsNi prévalence calibrée, ni vérité terrain. Validation humaine à venir. Le volet Gemma est une étiquette directe du modèle pour chaque travail de la base, lue sur la notice réduite au titre. Le volet Codex est un classifieur appris des 10 348 étiquettes directes de Codex et calibré sur les taux pondérés de l'échantillon; les champs sans appui suffisant ne portent aucun appel Codex. Le mode candidate est l'union des deux volets; le consensus est leur intersection. Ces sorties portent le statut machine_predicted_unvalidated et ne sont pas des étiquettes humaines.
Scores du classifieur distillé par catégorie (deux têtes)
| Catégorie | Codex | Gemma |
|---|---|---|
| Métarecherche | 0,007 | 0,114 |
| Méta-épidémiologie (sens strict) | 0,001 | 0,001 |
| Méta-épidémiologie (sens large) | 0,002 | 0,002 |
| Bibliométrie | 0,002 | 0,003 |
| Études des sciences et des technologies | 0,001 | 0,000 |
| Communication savante | 0,002 | 0,002 |
| Science ouverte | 0,002 | 0,001 |
| Intégrité de la recherche | 0,002 | 0,001 |
| Charge utile insuffisante (le modèle a refusé de juger) | 0,823 | 0,099 |
Scores machine (provisoires)
Les deux têtes enseignantes du modèle étudiant, lues sur ce travail. Un score ordonne la base pour la relecture; il n'affirme jamais une catégorie, et le statut de validation accompagne chaque rangée tel quel.
Scores de référence d'un modèle non mature (critères de maturité non atteints, 7 itérations). Un score ordonne; il n'affirme jamais une catégorie.
score_only:v0-immature-baseline · tel quel depuis la passe de notation : score_only signifie que le nombre peut ordonner les travaux, et qu'aucune étiquette de catégorie n'en découleClassification
machine, non validéePrédiction automatique; un appel candidat d’une seule source (Gemma direct ou Codex distillé), pas un consensus.
Le détail, modèle par modèle et score par score, se trouve en fin de page sous « Comment cette classification a été obtenue ».