Skip to content

请问如何得到Question-Level与KC Lecel (All-in-One)结果 #260

@KCAIED

Description

@KCAIED

您好,我基于最新版本的代码,并严格按照data_preprocess.py、wandb_dkt_train.py、wandb_predict.py的顺序进行操作,在assist2009数据集上训练DKT模型得到的dres结果如下:
{
"testauc": 0.8239183875382602,
"testacc": 0.7703965525589502,
"window_testauc": 0.8250769857245652,
"window_testacc": 0.7712436192653902,
"oriaucconcepts": 0.8239183875382602,
"oriauclate_mean": 0.8239179369660069,
"oriauclate_vote": 0.8239183875382602,
"oriauclate_all": 0.8239183875382602,
"oriaccconcepts": 0.7703965525589502,
"oriacclate_mean": 0.7704288319695283,
"oriacclate_vote": 0.7703965525589502,
"oriacclate_all": 0.7703965525589502,
"windowaucconcepts": 0.8250769857245652,
"windowauclate_mean": 0.8250764090558658,
"windowauclate_vote": 0.8250769857245652,
"windowauclate_all": 0.8250769857245652,
"windowaccconcepts": 0.7712436192653902,
"windowacclate_mean": 0.7712758248659441,
"windowacclate_vote": 0.7712436192653902,
"windowacclate_all": 0.7712436192653902
}
结合论文结果,这些数值大小看起来似乎对应的都是KC Level (One-by-One)模式下的结果,请问如何得到并查看Question-Level与KC Lecel分别在All-in-One模式下的结果以复现论文中的结果,或者需要在哪里进行额外设置,以实现上述不同的设定?

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions