Identifying Key Performance Indicators and Predicting the Result from Student Data

Student information systems hold a lot of information that can be mined for useful patterns. This work aims to build a prediction model to predict the result of students in 'C' Programming course by analyzing the factors that affect the performance of students. We applied feature selection techniques to select the most relevant academic and non-academic factors. The model is implemented using various classification algorithms and it is found that Nave Bayes classification model gives the highest accuracy of 82.4%.Decision tree based algorithm also showed considerable accuracy of 80.2%.The model was trained using 182 records from student dataset collected from the college with 20 attributes within the year 2008 to 2010. The model was validated using test records. It would predict the class label 'Result' as categorical value, Pass or Fail. Such a prediction model would help the faculty in early identification of 'at risk' students and thereby take timely and proactive measures to improve their performance.