Scribd
Upload
117 views

A Machine Learning Approach For Tracking and Predicting Student Performance in Degree Programs

The document proposes a machine learning approach for predicting student performance in degree programs. It develops a novel two-layer model using base predictors and ensemble predictors to make predictions based on students' evolving academic performance over multiple terms. It also uses latent factor models to discover the relevance of different courses, which is important for building effective base predictors. An evaluation on student data from UCLA shows the method outperforms benchmarks while maintaining educational interpretability.

Uploaded by

Imtiyaz Ali
Copyright
© © All Rights Reserved
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
117 views

A Machine Learning Approach For Tracking and Predicting Student Performance in Degree Programs

The document proposes a machine learning approach for predicting student performance in degree programs. It develops a novel two-layer model using base predictors and ensemble predictors to make predictions based on students' evolving academic performance over multiple terms. It also uses latent factor models to discover the relevance of different courses, which is important for building effective base predictors. An evaluation on student data from UCLA shows the method outperforms benchmarks while maintaining educational interpretability.

Uploaded by

Imtiyaz Ali
Copyright
© © All Rights Reserved
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
You are on page 1/ 2

A Machine Learning Approach for Tracking and Predicting Student

Performance in Degree Programs

Abstract:

Accurately predicting students’ future performance based on their ongoing


academic records is crucial for effectively carrying out necessary pedagogical
interventions to ensure students’ on-time and satisfactory graduation. Although
there is a rich literature on predicting student performance when solving problems
or studying for courses using data-driven approaches, predicting student
performance in completing degrees (e.g. college programs) is much less studied
and faces new challenges: (1) Students differ tremendously in terms of
backgrounds and selected courses; (2) Courses are not equally informative for
making accurate predictions; (3) Students’ evolving progress needs to be
incorporated into the prediction. In this paper, we develop a novel machine
learning method for predicting student performance in degree programs that is able
to address these key challenges. The proposed method has two major features.
First, a bilayered structure comprising of multiple base predictors and a cascade of
ensemble predictors is developed for making predictions based on students’
evolving performance states. Second, a data-driven approach based on latent factor
models and probabilistic matrix factorization is proposed to discover course
relevance, which is important for constructing efficient base predictors. Through
extensive simulations on an undergraduate student dataset collected over three
years at UCLA, we show that the proposed method achieves superior performance
to benchmark approaches.

Existing System:

First, students can differ tremendously in terms of backgrounds as well as their


chosen areas (majors, specializations), resulting in different selected courses as
well as course sequences. On the other hand, the same course can be taken by
students in different areas. Since predicting student performance in a particular
course relies on the student past performance in other courses, a key challenge for
training an effective predictor is how to handle heterogeneous student data due to
different areas and interests. In contrast, solving problems in ITSs often follow
routine steps which are the same for all students. Similarly, predictions of students’
performance in courses are often based on in-course assessments which are
designed to be the same for all students.

Disadvantage:

Students may take many courses but not all courses are equally informative for
predicting students’ future performance. Utilizing the student’s past performance
in all courses that he/she has completed not only increases complexity but also
introduces noise in the prediction, thereby degrading the prediction performance.
For instance, while it makes sense to consider a student’s grade in the course
“Linear Algebra” for predicting his/her grade in the course “Linear Optimization”,
the student’s grade in the course “Chemistry Lab” may have much weaker
predictive power. However, the course correlation is not always as obvious as in
this case. Therefore, discovering the underlying correlation among courses is of
great importance for making accurate performance predictions.

Proposed System:

We develop a novel algorithm for making predictions based on students’


progressive performance states. It adopts a bilayered structure comprising a base
predictor layer and an ensemble predictor layer. In the base layer, multiple base
predictors make local predictions given the snapshot of the student’s current
performance state in each academic term. In the ensemble layer, an ensemble
predictor issues a prediction of the student’s future performance by synthesizing
the local predictions results as well as the previous-term ensemble prediction. The
cascading of ensemble predictor over academic terms enables the incorporation of
students’ evolving progress into the prediction while keeping the complexity low.
We also derive a performance guarantee for our proposed algorithm.

Advantage:

We perform extensive simulation studies on an undergraduate student dataset


collected over three years across 1169 students at the Mechanical and Aerospace
Engineering department at UCLA. The results show that our proposed method is
able to significantly outperform benchmark methods while preserving educational
interpretability.

You might also like