Abstract—In this paper, a model for predicting students’
performance levels is proposed which employs three machine
learning algorithms: instance-based learning Classifier,
Decision Tree and Naïve Bayes. In addition, three decision
schemes were used to combine results of the machine learning
techniques in different ways to investigate if better classification
performance could be achieved. The experiment consists of two
phases that are testing and training. These phases are
conducted at three steps which correspond to different stages in
the semester. At each step the number of attributes in the
dataset has been increased and all attributes were included at
final stage. The important characteristic of the dataset was that
it only contains time-varying attributes rather than
time-invariant attributes such as gender or age. This type of
dataset has helped to learn to what extend time-invariant data
has significant effect on prediction accuracy. The experiment
results were evaluated in terms of overall accuracy, sensitivity
and precision. Results are discussed compared to results
reported in the relevant literature.
Index Terms—Machine learning, online learning, students'
performance prediction.
Erkan Er is with the Department of Information Systems, Informatics
Institute, Middle East Technical University, Ankara, Turkey (e-mail: er@
metu.edu.tr).
Cite:Erkan Er, "Identifying At-Risk Students Using Machine Learning Techniques: A Case Study with IS 100," International Journal of Machine Learning and Computing vol.2, no. 4, pp. 页码, 2012.