Thursday, October 19, 2017
   
Text Size
Login

Journal of Information and Communication Technology (JICT) Vol. 7, 2008

MULTILEVEL KOHONEN NETWORK LEARNING FOR CLUSTERING PROBLEMS
Siti Mariyam Shamsuddin, Anazida Zainal and Norfadzila Mohd Yusof
Soft Computing Research Group
Universiti Teknologi Malaysia
 
ABSTRACT FULL TEXT
Clustering is the procedure of recognising classes of patterns that occur in the environment and assigning each pattern to its relevant class. Unlike classical statistical methods, self-organising map (SOM) does not require any prior knowledge about the statistical distribution of the patterns in the environment. In this study, an alternative classification of self-organising neural networks, known as multilevel learning, was proposed to solve the task of pattern separation. The performance of standard SOM and multilevel SOM were evaluated with different distance or dissimilarity measures in retrieving similarity between patterns. The purpose of this analysis was to evaluate the quality of map produced by SOM learning using different distance measures in representing a given dataset. Based on the results obtained from both SOM methods, predictions can be made for the unknown samples. The results showed that multilevel SOM learning gives better classification rate for small and medium scale datasets, but not for large scale dataset.
 
Keywords: Classification, Patterns, Self-organising map, SOM, Multilevel learning, Distance (or dissimilarity) measure, Predictions, Computational times, Classification rate.
 

 
CLASSIFICATION OF MALAYSIAN VOWELS USING FORMANT BASED FEATURES
S. A. Mohd Yusof, Paulraj M, S. Yaacob
School of Mechatronics Engineering, Universiti Malaysia Perlis,
01000 Kangar, Perlis, Malaysia
shahrulazmi@uum.edu.my
paul@unimap.edu.my
s.yaacob@unimap.edu.my
 
ABSTRACT FULL TEXT
Automatic speech recognition (ASR) has made great strides with the development of digital signal processing hardware and software, especially using English as the language of choice. Despite of all these advances, machines cannot match the performance of their human counterparts in terms of accuracy and speed, especially in case of speaker independent speech recognition. In this paper, a new feature based on formant is presented and evaluated on Malaysian spoken vowels. These features were classified and used to identify vowels recorded from 80 Malaysian speakers. A back propagation neural network (BPNN) model was developed to classify the vowels. Six formant features were evaluated, which were the first three formant frequencies and the distances between each of them. Results, showed that overall vowel classification rate of these three formant combinations are comparatively the same but differs in terms of individual vowel classification.
 

 
NON-TAYLOR SERIES BASED POSITIONING METHOD FOR LOCATION BASED SERVICES
Tan Kim Geok, Ahmed Wasif Reza and San Choy Liew
Faculty of Engineering & Technology
Multimedia University
Jalan Ayer Keroh Lama, Bukit Beruang
75450 Melaka, Malaysia
awreza98@yahoo.com
 
ABSTRACT FULL TEXT
Location Based Services (LBS) has gained increasing popularity in major cities. Due to blocking from man-made structures, the existing Global Positioning System (GPS) could not satisfy LBS applications, especially in street canyon and indoor surroundings. This has lead to the development of Assisted GPS (A-GPS) which can provide better service availability and accuracy gain. In the conventional positioning method, Taylor series expansion is applied to solve non-linear distance equations. This method requires an initial estimation of A-GPS receiver’s position. This paper investigates the positioning method for LBS based on hybrid E-OTD/GNSS. The proposed positioning method is non-Taylor series based. Therefore, it involves less complicated mathematical expansion and substitution. A flexible LBS positioning tool is developed which can generate position information in convenient way. It supports both Taylor series and non-Taylor series based positioning methods. The obtained results showed that the proposed non-Taylor series based positioning method can achieve better positioning accuracy.
Keywords: Location based services, Taylor series, Positioning system.
 

 
THE COMMERCIALISATION OF KNOWLEDGE MANAGEMENT PRACTICES TO K-BASED DEVELOPMENT IN MALAYSIA
Raja Suzana Raja Kasim
Universiti Teknologi MARA, Shah Alam, Selangor, Malaysia
rsuzana@salam.uitm.edu.my
 
ABSTRACT FULL TEXT
This paper studies the connection between the commercialization of knowledge management practices and its contributions toward the knowledge-based development in Malaysia. Theoretical relations in this paper were tested through an empirical study carried out among public-listed organisations in the industrial products, consumer products, and service industries in Malaysia. The findings of this paper revealed that it is important for corporations to focus on knowledge management in the development of the organisation’s corporate strategy. Empirical evidence supported the view that firms with knowledge management capabilities and understanding will utilise these capabilities to drive their quality strategy, thus performing better than those that do not. In addition, the finding serves as an initial move among private sector business toward a knowledge-based country. This paper concludes that these practices have a positive incidence not only on the performance of the firm but it also gives some initial contributions toward the development of Malaysia as a knowledge-based country.
 
Keywords: Knowledge-based development, Corporate strategies, Firm performance, Public-listed organisations, Malaysia.
 

 
ENHANCED MAINTENANCE PROBLEM RECOGNITION TECHNIQUES AND ITS APPLICATION TO PALM OIL MILLS
Basari, A.S.H., Herman, N.S. and Asmai, S.A.
Faculty of Information and Communication Technology
Universiti Teknikal Malaysia Melaka
Karung Berkunci 1200, 75450 Ayer Keroh, Melaka, Malaysia
abdsamad@utem.edu.my
nsuryana@utem.edu.my
 
ABSTRACT FULL TEXT
This paper discusses the application of enhanced maintenance problem recognition techniques. The main contribution of this study is the proposed combined techniques, namely snapshot model, failure mode, effect and criticality analysis (FMECA), Pareto analysis, and decision analysis by using information technology (IT). The snapshot model is part of the maintenance modelling technique while FMECA, Pareto analysis, and decision analysis are part of maintenance reliability techniques. Each of the techniques and the proposed combined techniques is explained. The case study used for this enhanced technique was the palm oil mills maintenance problem. The result and possible further enhancement is also discussed.
 
Keywords: Maintenance problem recognition, Snapshot model
.
 
PERFORMANCE EVALUATION AND INDICES OF CYBER CAFÉ BUSINESS: A FACTOR ANALYTIC APPROACH
Oluwadare Samuel Adebayo
Department of Computer Science,
Federal University of Technology,
P.M.B. 704, Akure
 
ABSTRACT FULL TEXT
The advent of the World Wide Web has opened up a new vista of opportunities for investment in the Information Communication Technology (ICT) industry. In recent years, the cyber café business has attracted considerable investment, providing job opportunities for many people. On the other hand, it has also been observed that some hitherto vibrant cyber cafes are also closing down due to the fact that they could not break even. This study, is therefore, aimed at identifying the factors/indices that could enhance the performance of cyber cafes. A total of 250 users of cyber cafes in Akure, Ondo State, Nigeria were randomly selected and a structured questionnaire was administered on them. Data collected was analysed using factor analysis by principal component. The result revealed that four factors, namely quality of hardware, speed of processing, cost, and reliability of service are the major factors that influence users’ patronage. It is therefore recommended that investors and would-be-investors in the cyber cafe business should pay attention to these facors in order to enhance the success of their investment.
 
Keywords: Performance evaluation, Indices, Cyber cafe, Factor analysis.

Universiti Utara Malaysia Press
 Universiti Utara Malaysia, 06010 UUM Sintok
Kedah Darul Aman, MALAYSIA
Phone: +604-928 4816, Fax : +604-928 4792

All Right Reserved. Copyright © 2010, Universiti Utara Malaysia Press