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

Determining Suitable Program For SPM Holder Using Neural Network Approach

Noraisah, Sudin (2002) Determining Suitable Program For SPM Holder Using Neural Network Approach. Masters thesis, Universiti Utara Malaysia.

[img] PDF
NORAISAH_SUDIN.pdf
Restricted to Registered users only

Download (10MB)
[img]
Preview
PDF
1.NORAISAH_SUDIN.pdf

Download (1MB) | Preview

Abstract

This study involves the development of a neural network (NN) prediction model proposed for the educational spanding spesifically in determining suitable program for SPM Holder . This module used for self-assesment prediction in web-based environment in order to get more user participation. The basic architecture are multilayer feedforward networks, trained using the Backpropagation algorithm. The evaluation using 302 data sets showed that the developed architecture is very useful for high dimensional input vectors. The sample data has been collected from 4 study programs in four Ministry of Education Matriculation Centre, from UiTM and Shahputra College. The data was then trained using the proposed model. The findings show that the most suitable predictive model comprises of 35 nodes in input-layer; three nodes in hidden layer and one node in output-layer. The generalization performance of the selected model is 83.33%. This methodology should be able to provide us with some new insights into the type of patterns that exist in educational data. Therefore, NN has a great potential in supporting the policy development for the current education.

Item Type: Thesis (Masters)
Uncontrolled Keywords: Neural Network, Multilayer Feedforward Network, Backpropagation Algorithm, SPM Holder
Subjects: Q Science > QA Mathematics > QA76 Computer software
Divisions: Faculty and School System > Sekolah Siswazah
Depositing User: Mrs Hafiza Mohd Akhir
Date Deposited: 04 Nov 2009 07:10
Last Modified: 24 Jul 2013 12:08
URI: http://etd.uum.edu.my/id/eprint/684

Actions (login required)

View Item View Item