Please use this identifier to cite or link to this item:
|Title:||A study on Implementation of classification techniques to predict students’ results for Institutional Analysis|
|Authors:||Gorikhan, Naseem Akhtar|
|Publisher:||The British University in Dubai (BUiD)|
|Abstract:||This thesis presents an implementation of classification techniques for a Vocational Institutional analysis. The institute is known as IAT. The classification techniques used were decision tree, knn, logistic regression, support vector and neural network and it was found that the decision tree proved out to be accurate prediction model for institute’s analysis of students’ results. Based on the prediction, teachers in institution worked on weak students to improve their performance. After final exam result declaration, the results were compared with previous year results and it was found that the classification technique helped the institution to increase the overall passing average in computer science course. Moreover, the prediction analysis was applied for newly enrolled students. An educational institution must always have an estimated previous knowledge of enrolled students to predict their performance in future academics. This assists many decision makers in educational field to identify talented students and to focus on low achievers in order to improve their grades. This thesis emphasizes on data mining tasks that will predict the academic performance of students in CS (Computer Science) exam by considering their grades in math and science from previous exam. The prediction models are developed using classification techniques such as decision tree, knn, logistic regression, support vector and neural networks. The outcome of these models is to predict the number of students who were likely to pass or fail. The results were given to teachers and steps were taken to improve the academic performance of the weak/ fail students. After final examination, CS exam results of year 2015 were fed in the system for analysis and then compared with the previous year results (2014). The comparative analysis of results states that the prediction has helped the weaker students to improve their marks in CS exam which has eventually lead to increased overall passing average of the CS course. In this thesis, the analysis was done using classification models with and without math and science marks of previous exam, the models are then compared to select the prediction model that produced highest accuracy, which helped the institute to identify the students likely to fail, and work on their academics accordingly in order to achieve better results.|
|Description:||DISSERTATION WITH DISTINCTION|
|Appears in Collections:||Dissertations for Informatics (Knowledge and Data Management)|
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.