Please use this identifier to cite or link to this item:
https://ptsldigital.ukm.my/jspui/handle/123456789/513357Full metadata record
| DC Field | Value | Language |
|---|---|---|
| dc.contributor.advisor | Zarina Shukur, Prof. Dr. | - |
| dc.contributor.author | Mansour Esmaeilpour (P47843) | - |
| dc.date.accessioned | 2023-10-16T04:35:50Z | - |
| dc.date.available | 2023-10-16T04:35:50Z | - |
| dc.date.issued | 2012-09-06 | - |
| dc.identifier.other | ukmvital:120174 | - |
| dc.identifier.uri | https://ptsldigital.ukm.my/jspui/handle/123456789/513357 | - |
| dc.description | Knowledge Analysis is an approach to analyze problems, issues or events through a knowledge perspective in order to understand them at a deeper level and form new insights about them. In this work, a family of learning automata that is Cellular Learning automata (CLA) and Distributed Learning automata (DLA) are blended together in order to obtain a better processing time and accuracy in analyzing knowledge. A technology is developed in order to support the overall process of the analysis using this hybrid method. The name of this technology is CADAR. Individually, CLA and DLA are proven powerful methods for knowledge analyzing such as clustering, classification and frequent pattern mining. However, based on the literature, no work has been done to demonstrate the ability of combining CLA and DLA in analyzing knowledge. It is hypothesized that by combining both methods, better result especially in accuracy and run time can be obtained. As for CADAR, its aim is to provide a favorable environment for knowledge analyst to use the proposed method. Conceptually, learning automata (LA) is composed of two parts; stochastic automata with a number of limited actions and a stochastic environment. Each action selected by potential environment is assessed and answer is given to a learning automata. LA uses this answer in order to select its action for the next stage. As for CLA, it is composed of two models; learning automata and cellular automata. The learning automata is assigned to every cell in cellular automata. It is suitable to be used for systems that can be represented by a cellular model where each cell's behavior is based on its neighbor's behavior and past experience. Whilst DLA is a network of learning automata which cooperates with each other for solving problem. In this research, in order to acquire better result, it is combined CLA and DLA for knowledge analysis. In this proposed method, each node of the DLA is one CLA. The node is created dynamically based on the data. The gist of this hybrid method is the ability to capture the amount of strong relationship reinforcement. Algorithm of the proposed method is the main engine of CADAR. It is implemented by using dynamic arrays where its size can change during runtime. By having dynamic arrays, it increases the efficiency of processing time. Besides the algorithm of the methods, CADAR also consider usability aspects so that the knowledge analyst can have better use of the technology. The performance of the proposed method compared to well known methods have been investigated by using an experiment. Whilst the performance of the technology itself compared to WEKA tool is done by using a simple usability study. The result demonstrates that the proposed method can works on all of kind of datasets and it is suitable and acceptable from the aspect of the run time and excellent in accuracy compared to other methods. As for the technology, the result shows that CADAR wins over WEKA on four features that are: data cleaning and discretization function, graphical report of data clustering, frequent pattern mining function, and rule generator.,Certification of Master's / Doctoral Thesis" is not available | - |
| dc.language.iso | eng | - |
| dc.publisher | UKM, Bangi | - |
| dc.relation | Faculty of Information Science and Technology / Fakulti Teknologi dan Sains Maklumat | - |
| dc.rights | UKM | - |
| dc.subject | Cellular automata | - |
| dc.subject | Universiti Kebangsaan Malaysia -- Dissertations | - |
| dc.subject | Dissertations, Academic -- Malaysia | - |
| dc.title | CADAR: knowledge analyzer using cellular learning automata and distributed cellular learning automata | - |
| dc.type | Theses | - |
| dc.format.pages | 128 | - |
| dc.identifier.callno | QA267.5.C45E846 2012 3 tesis | - |
| dc.identifier.barcode | 004072(2019) | - |
| Appears in Collections: | Faculty of Information Science and Technology / Fakulti Teknologi dan Sains Maklumat | |
Files in This Item:
| File | Description | Size | Format | |
|---|---|---|---|---|
| ukmvital_120174+SOURCE1+SOURCE1.0.PDF Restricted Access | 331.25 kB | Adobe PDF |  View/Open |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.