Please use this identifier to cite or link to this item:
https://ptsldigital.ukm.my/jspui/handle/123456789/476205Full metadata record
| DC Field | Value | Language |
|---|---|---|
| dc.contributor.advisor | Azuraliza Abu Bakar, Dr. | |
| dc.contributor.author | Faiza Abdulsalam M. Mansour (P49119) | |
| dc.date.accessioned | 2023-10-06T09:14:40Z | - |
| dc.date.available | 2023-10-06T09:14:40Z | - |
| dc.date.issued | 2011-04-15 | |
| dc.identifier.other | ukmvital:75275 | |
| dc.identifier.uri | https://ptsldigital.ukm.my/jspui/handle/123456789/476205 | - |
| dc.description | This study aims to employ the Artificial Bee Colony (ABC) algorithm for cluster based deviation detection. Artificial Bee Colony (ABC) algorithm is one of the most newly studied algorithms based on the intelligent foraging behavior of honey bee swarms. Currently, ABC algorithm was developed to solve clustering problems and revealed very promising results in terms of time required for processing and the solution quality. However, to date none of this research looks in the different aspect of employing ABC for deviation detection. Deviation detection is a data mining task with wide variety of application domains, including fraud detection, criminal activities in e-commerce, detecting suspicious activities, and computer network intrusion detection. Several cluster based deviation detection techniques have been developed. Most of these techniques depend on the key assumption that normal objects belong to large and dense clusters, while deviations belong to very small clusters. The object that significantly deviates from the other object may be identified as deviations and are not assigned to any cluster. In this study, we propose a modification of the existing ABC clustering algorithm for the purpose of deviation detection. Two steps are involved; i) the development of ABC clustering algorithm and ii) modifying the ABC clustering for deviation detection. The proposed method intends to cluster the dataset into many clusters and detecting data points that do not belong to any cluster and assign them as deviating objects. Outlier factor has been used to identify top n outliers that deviate from the dataset. The proposed algorithm was tested upon the UCI benchmark datasets. Three of these datasets undergo the data preparation process for the purpose of deviation detection. The performances of the deviation detection task are measured in Detection Rate (DR) and False Alarm Rate (FR) measurements. Two cluster based deviation detection methods are used for comparison and evaluating the proposed method. Experimental results have shown that ABC deviation detection algorithm has achieved its performance with comparable results in terms of detecting deviations.,Master of Information Technology | |
| 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 | Computer algorithms | |
| dc.title | Clustered based deviation detection task using artificial bee colony (ABC) algorithm | |
| dc.type | theses | |
| dc.format.pages | 101 | |
| dc.identifier.callno | QA76.9.A43M345 2011 tesis | |
| dc.identifier.barcode | 000801 | |
| Appears in Collections: | Faculty of Information Science and Technology / Fakulti Teknologi dan Sains Maklumat | |
Files in This Item:
| File | Description | Size | Format | |
|---|---|---|---|---|
| ukmvital_75275+Source01+Source010.PDF Restricted Access | 1.37 MB | Adobe PDF |  View/Open |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.