A stacking ensemble of LG, SVM, NB, KNN and XGB achieved the best fine-tuning F1-score (0.88), outperforming individual models such as SVM (0.86)¶

Claim · All claims
Evidence · 1 study · design r2 · q2
Currency · newest study 2026 · not yet checked against its sources · not reviewed by a person

Subclaims¶

q2 i? Among 120 evaluated stack configurations, the best-performing stack (LG, SVM, NB, KNN, XGB) achieved an F1-score of 0.88 during fine-tuning, exceeding SVM's second-highest 0.86. → Josmario Albuquerque 2026

Evidence¶

Josmario Albuquerque 2026¶

Josmario Albuquerque, Bart Rienties, Martin Hlosta, Wayne Holmes. (2026). Learning Analytics to Uncover Ethnic Bias in Educational Texts: An Ensemble Learning Approach. Journal of Learning Analytics 13(1). https://doi.org/10.18608/jla.2026.8905

q2 · i? · design · r2

Fine-tuning experiment evaluating 120 stacking combinations by F1-score during training/validation on the OER dataset; the stack "outperformed individual models such as SVM" at this stage, though on the test set STK's F1 was 0.69.

"the best-performing stack achieved an F1-score of 0.88. This stack included LG, SVM, NB, KNN, and XGB models and outperformed individual models such as SVM, which achieved the second-highest F1-score of 0.86."

Discussion¶

Related Claims¶

• SVM and random forest classifiers showed consistent performance in classifying ethnic bias, with F1-scores of 0.71 and 0.70 on the test set — related
• The naive Bayes model achieved the highest precision (0.75) for ethnic bias detection on the test set — related
• Both pedagogy-expertise-weighted and unanimous-vote ensembles frequently worsen LLM alignment with student learning — related
• Ensemble modeling of multiple LLM scorers yields performance gains in automated divergent thinking scoring — a narrower finding that bears on this claim
• A Soft-Voting Ensemble achieves the best behavioral-dataset attrition prediction (CV F1 = 0.737, test recall = 0.769) — a broader claim this one bears on
• Random forest algorithms yielded the best classification performance in K–8 MMLA studies comparing multiple machine learning models — related
• Machine learning methods outperformed logistic regression in classification accuracy across reviewed studies — related
• k-nearest neighbours achieves the highest recall of the tested models, making it better suited when identifying which students know a topic matters — related
• Rank-averaged ensembling stabilized predictions but did not outperform the single base model — related