Please use this identifier to cite or link to this item:
https://ptsldigital.ukm.my/jspui/handle/123456789/476563Full metadata record
| DC Field | Value | Language |
|---|---|---|
| dc.contributor.advisor | Shahnorbanun Sahran, Prof. Dr. | - |
| dc.contributor.author | Hamid Yadollahi (P65642) | - |
| dc.date.accessioned | 2023-10-06T09:21:09Z | - |
| dc.date.available | 2023-10-06T09:21:09Z | - |
| dc.date.issued | 2017-02-29 | - |
| dc.identifier.other | ukmvital:120177 | - |
| dc.identifier.uri | https://ptsldigital.ukm.my/jspui/handle/123456789/476563 | - |
| dc.description | In face recognition the feature extraction process in a supervised learning based technique usually creates a large set of features leading to a substantial computation cost in the course of classification. Furthermore, the availability of enormous amounts of features represents a challenge to classification problems. Feature selection is a worldwide optimization problem in machine learning in order to select the optimum and instructive features from an original set by excluding irrelevant and redundant features which Decreases the quantity of features, eliminates irrelevant, noisy and dismissed data, and results in acceptable recognition accuracy. In this research feature selection optimization algorithms will be implemented on embedded feature selection method to challenge the classification accuracy and performance of the model comparing with previous works. Neural network based backpropagation algorithm (BPNN) for classification technique, Discrete Wavelet Transform (DWT) and Pseudo Zernike Moment Invariant (PZMI) for features extraction method are applied in this research. This research presents Ant Colony Optimization (ACO) and Bat Algorithm (BAT) as our proposed feature selection optimization algorithms. ACO is inspired of ant's social behavior in their search for the shortest paths to food sources and BAT is the idea of echolocation behavior of bats which find their prey by sense distances. In the proposed algorithms, classifier performance and the length of selected feature vector are adopted as heuristic information. Therefore, we can select the optimum feature subset without the prior knowledge of features. Simulation results of the face recognition model on ORL face database present the superiority of the proposed algorithms. PZMI facial feature extraction method is slightly better than DWT feature but the time for PZMI is higher than DWT. Furthermore, AS-rank could provide less feature size for higher recognition rate which shows its efficiency. A part from the processing time, it seems the best feature selection methods are AS-rank, Ant Colony System (ACS) and BAT respectively. Also, it can be concluded that whenever the size of population is decreased, the recognition rate is increased. Furthermore, increasing the feature size can improve the recognition rate but in special point, increasing the feature subset cannot seriously increase the recognition rate which is detected by feature selection method. The experiments indicate that the proposed face recognition system with selected features is more practical and efficient when compared with others.,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 | Wavelets (Mathematics) | - |
| dc.subject | Ant algorithms | - |
| dc.subject | Image processing -- Digital techniques | - |
| dc.subject | Human face recognition (Computer science) | - |
| dc.subject | Universiti Kebangsaan Malaysia-- Dissertations | - |
| dc.subject | Dissertations, Academic -- Malaysia | - |
| dc.title | Feature selection optimization using ACO and Bat Algorithms in face recognition | - |
| dc.type | theses | - |
| dc.format.pages | 108 | - |
| dc.identifier.callno | TA1653.Y333 2017 3 tesis | - |
| dc.identifier.barcode | 004075(2019) | - |
| Appears in Collections: | Faculty of Information Science and Technology / Fakulti Teknologi dan Sains Maklumat | |
Files in This Item:
| File | Description | Size | Format | |
|---|---|---|---|---|
| ukmvital_120177+SOURCE1+SOURCE1.0.PDF Restricted Access | 372.67 kB | Adobe PDF |  View/Open |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.