Predicting Students performance using k-means algorithm

A system is developed for analyzing students results using cluster analysis and uses standard statistical algorithms to arrange their scores data according to the level of their performance. We take into account the GPA and scores of a student to predict their overall academic performance. This performance is targeted by the faculty members to improve student learning and academic performance. First the scores of different students are the dataset. Initialize k clusters by assuming k clusters or by random sampling. Then, each record is assigned to the nearest cluster by smallest euclidean distance measurement. Reassign and recalculate the mean of all clusters. Iteration is done until a precise value is obtained. An algorithm using the codes are developed for the whole process of assigning the k-clusters to finding the precise values  are developed which can be found in the reference link. Results for k=3, k=4 and k=5 are displayed in paper provided in reference link. Finally at k=5, we have cluster 1 of size 19 with overall performance of 49.85, cluster 2 of size 17 with overall performance of 60.97, cluster 3 of size 9 with overall performance of 43.65, cluster 4 of size 14 with performance of 64.93, cluster 5 of size 20 with a performance of 55.79.  This is used to evaluate the student data to describe different dependencies between the attributes and the student status. This clustering algorithm serves as a good benchmark to monitor the progression of students’ performance in higher institution. It also enhances the decision making by academic planners to monitor the candidates’ performance semester by semester by improving on the future academic results in the subsequence academic session.

Reference: http://arxiv.org/ftp/arxiv/papers/1002/1002.2425.pdf