Models trained with combined DKT/DKT+ knowledge-state and student-profile features generally outperform models trained with student profile alone on STEM/non-STEM job prediction.¶

Claim · All claims
Evidence · 1 study · associational r1 · q2
Currency · newest study 2018 · not yet checked against its sources · not reviewed by a person

Subclaims¶

q2 i? Test results show models trained with DKT or DKT+ knowledge states, alone or combined with the student profile, generally perform better than models trained with the student profile alone. → Yeung 2018

Evidence¶

Yeung 2018¶

Yeung, C.-K., Lin, Z., Yang, K., & Yeung, D.-Y. (2018). Incorporating Features Learned by an Enhanced Deep Knowledge Tracing Model for STEM/Non-STEM Job Prediction. arXiv preprint arXiv:1806.03256. https://arxiv.org/abs/1806.03256

q2 · i? · associational · r1

Cross-validation experiment comparing five feature sets (student profile, DKT, DKT+, and combined) across GBDT, LDA, LR and SVM classifiers, reported in Section 5.1 and Tables 2–3. The authors state that "the models trained with the DKT or DKT+ knowledge state, as well as the combined features, result in a higher performance than the models trained with the student profile alone."

"Generally, the models trained with the DKT or DKT+ knowledge state, as well as the combined features, result in a higher performance than the models trained with the student profile alone."

Discussion¶

Related Claims¶

• LDA projection visualizations show knowledge-state features separate STEM and non-STEM classes better than student-profile features. — related
• Recursive feature elimination mitigates overfitting and improves STEM/non-STEM prediction performance, with LR-DKT&SP performing best after RFE. — related
• DKT+ knowledge states discriminate STEM from non-STEM students better than DKT knowledge states for the majority of mathematical skills. — related
• STEM students show a higher mastery level and a higher learning gain in mathematics than non-STEM students. — related