Nowadays, especially with the recent rise of online educational platforms and e-learning,
there are more opportunities to take advantage of and help improve the educational sector.
One such opportunity is the prediction of student performance, or as it is called in the
Educational Data Mining community Knowledge Tracing, using a machine learning
model, to optimize each student’s learning process and help them improve their performance by deriving conclusions from the model’s results. The results gained from the
model can help both teacher and students. Teachers gain the ability to pinpoint their student’s weak points and potentially emphasize on assisting them combat those weaknesses,
but also track their progress on these subjects. Students on the other hand can work on
their own weaker subjects and go through a self-improvement process.
The goal of this project is to present one such model with credible results which could
help improve the educational sector. The model follows a time delay architecture and uses
various deep learning methodologies and layers to achieve its purpose. It has been used
on three (3) different datasets with no architectural or parameter changes from one dataset
to another.
Collections
Show Collections