•   When: Wednesday, May 01, 2019 from 01:30 PM to 03:30 PM
•   Speakers: Zhiyun Ren
•   Location: ENGR 4201
•   Export to iCal

The past decade has seen a growth in the development and deployment of data mining technologies utilized in educational environments for assisting students in selecting courses, acquiring feedback and improving performance based on past academic performance and behaviors. Grade prediction methods seek to estimate a grade that a student may achieve in a course/task that she may take in the future (e.g., next term, next assessment). Accurate and timely prediction of students’ academic grades holds the promise for better student degree planning, personalized advising and automated interventions to ensure that students stay on track in their chosen degree program and graduate on time.

The existing grade prediction methods are mainly based on matrix factorization approaches, and always overlook important factors that could greatly influence student’s performance. In our previous work, we have developed several methods on grade prediction to tackle this problem for both traditional university environment and online learning systems. Specifically, we consider that a student’s knowledge is continuously being enriched while taking a sequence of courses and propose a model named Matrix Factorization with Temporal Course-wise Influence. Furthermore, we substitute student’s latent factors with accumulated knowledge of a sequence of courses taken by the student, jointly with the grade for each course. And we incorporate course instructor and student academic level effects along with student global latent factor to complete grade prediction. We name this model Additive Latent Effect. Moreover, we present next-term grade prediction models based on students’ cumulative knowledge and co-taken courses. The proposed models are based on a matrix factorization framework and incorporate a co-taken course interaction function to learn the influence from the co-taken courses on the target course. The co-taken course interaction function is formed by a neural network, which takes the knowledge difference between the co-taken courses and the target course as input, and outputs an influence value that will be used to predict students’ grades on the target course. Finally, we present a deep learning based recommender system approach called Neural Collaborative Filtering (NCF) for predicting the grade a student will earn in a course that he/she plans to take in the next-term. The deep learning inspired approach provides added flexibility in learning the latent spaces in comparison to MF approaches. The proposed approach also incorporates instructor information besides student and course information. In addition, we also apply a Personalized Linear Multi-Regression model to predict student’s performance on online education environment, i.e., Massive Open Online Courses (MOOCs), and gain great results.