UUM ETD | Universiti Utara Malaysian Electronic Theses and Dissertation
FAQs | Feedback | Search Tips | Sitemap

Neural Network Model for Predicting Students' Performance

Lee, Teow Wak (2003) Neural Network Model for Predicting Students' Performance. Masters thesis, Universiti Utara Malaysia.

[img] PDF
Restricted to Registered users only

Download (9MB)


Data mining techniques have been used to search for patterns or trends in data that may help to explain certain outcomes. Neural networks model of multilayer perception with back-propogation has also been successfully used in prediction and classification in fields such as medicine, business and education. While neural networks are Musk boxes that process the inputs and genarate the outputs, statistical methods can help to establish relationship between variables and give a meaningful explanation to certain outcomes. This study uses all the above techniques and tools to explain and predict the performance of students in the SPM examination. For the purpose of this study data was collected from the year 2002 Form 5 students of Sekolah Menengah Teknik Alor Star. Data mining technique and statistical methods are used to identify suitable inputs into a buck propagation neural network. The findings indicate that the highest generalization performance of the network is 65.6% for a network of six input units, six hidden units, and one output unit. Statistical analysis reveals that the stronges linear relationship exists between the average SPM grade and the average SPM trial examination marks. It is recommended that further study be carried out to predict the profile of the students.

Item Type: Thesis (Masters)
Uncontrolled Keywords: Neural Networks, Academic Achievements, Sijil Pelajaran Malaysia (SPM) Examination
Subjects: Q Science > QA Mathematics > QA71-90 Instruments and machines
Divisions: Faculty and School System > Sekolah Siswazah
Depositing User: Mrs Shamsiah Mohd Shariff
Date Deposited: 30 Dec 2009 01:07
Last Modified: 24 Jul 2013 12:10
URI: http://etd.uum.edu.my/id/eprint/1024

Actions (login required)

View Item View Item