1 paper
Houru Jiang, Zixi Wang, Tengteng Cheng +6
Knowledge tracing (KT) models are predominantly evaluated using aggregate metrics such as area under the curve (AUC) and accuracy. However, these global scores obscure where the re…