Abstract
Nowadays, clustering is one of the popular technique to grouping data to make them ease to interpret the result. Some researcher even used clustering to predict weather, natural disaster and many others. This prediction can be made by cluster the new data with the old data and observe where the data belongs. Even a normal people using clustering to grouping their data. But all the clustering tool available are not suitable enough for a normal people that does not have expert knowledge in this field. In order to make the normal people can cluster their data easily, this project aims is to develop a web-based clustering tool that can be used by all peoples. This project will use fuzzy k-means clustering algorithm to cluster the data because it is easy to implement and have many advantages. Moreover, silhouette method has been implemented to check the score of the clustering result. This project is developed by using Rapid Application Development methodology as it is the model that is most suitable for developing this web tools. This methodology consist of Requirement and Planning, Design, Construction, Testing and Implementation. In Requirement and Planning phase, the problem statement, objective and scope has been describe briefly. The study on related work and comparison of algorithm also has been done in this phase. In Design phase, the use case diagram, whole system flowchart and subsystem flowchart has been constructed to assist the development of this web tool. On the Construction phase, the development of the prototype has been started. All the algorithm for the engine has been developed by using Java script language. Lastly, on the Testing phase two type of data has been used to test this web tool result. Iris data set has been taken from UCO machine learning to assist in testing phase. The validity of the result for this web tool will be determined by the average score of the silhouette method. Although the web tool has some limitation which it can hold back the user experience on this web tool, but there are always room for improvement.
Metadata
Item Type: | Thesis (Degree) |
---|---|
Creators: | Creators Email / ID Num. Zulkifly, Ahmad Zuladzlan 2016537615 |
Contributors: | Contribution Name Email / ID Num. Thesis advisor Seman, Ali UNSPECIFIED |
Subjects: | Q Science > QA Mathematics > Instruments and machines > Electronic Computers. Computer Science |
Divisions: | Universiti Teknologi MARA, Shah Alam > Faculty of Computer and Mathematical Sciences |
Programme: | Bachelor of Computer Sciences (Hons.) |
Keywords: | Web-based, clustering application, algorithm |
Date: | 2019 |
URI: | https://ir.uitm.edu.my/id/eprint/110709 |
Download
![[thumbnail of 110709.pdf]](https://ir.uitm.edu.my/style/images/fileicons/text.png)
110709.pdf
Download (145kB)
Digital Copy

Physical Copy
ID Number
110709
Indexing

