Should an separate instance of model be created for each student in actual use?

lccasagrande / Deep-Knowledge-Tracing

An implementation of the Deep Knowledge Tracing (DKT) using Tensorflow 2.0

MIT License

95 stars 38 forks source link

Should an separate instance of model be created for each student in actual use? #6

Open zwhe99 opened 4 years ago

zwhe99 commented 4 years ago

Thanks for your python implmentation !

I have the following question： Should an separate instance of model be created for each student in actual use? Or, is the interaction sequence of all students going through a single model in order (according to timestamp)？

lccasagrande commented 3 years ago

The idea here is to find the probability of getting a question right based on past answers. Therefore, you must include data from all students.

In other words, how would you know that mastering skill "X" influences the probability of mastering skill "Y" without considering every student in the database? Some students may guess the answer, which does not mean that they mastered the required skill to get the question right.