Publication: Improved landslide susceptibility map prediction of penang island using hybrid artificial bee colony-artificial fish swarm models
| datacite.subject.fos | oecd::Engineering and technology::Electrical engineering, Electronic engineering, Information engineering | |
| dc.contributor.author | Ilyas Ahmad Huqqani | |
| dc.date.accessioned | 2025-05-08T03:01:21Z | |
| dc.date.available | 2025-05-08T03:01:21Z | |
| dc.date.issued | 2023-10-01 | |
| dc.description.abstract | Penang Island is one of the landslide-prone areas in Malaysia, with high landslide occurrences causing damage to farmland, non-cultivated lands, resources, and even threats to human lives. To mitigate the landslide hazard and minimize property damage and loss of life, landslide susceptibility mapping is required. Existing algorithms for landslide susceptibility analysis exhibit some limitations such as local minima, overfitting, and dimension danger. Besides, meta-heuristic algorithms including artificial bee colony (ABC) and artificial fish swarm (AFS) have proven to be capable in solving optimization problems in many applications. Therefore, an innovative approach is proposed for landslide susceptibility mapping utilizing a hybrid of artificial bee colony-artificial fish swarm (ABC-AFS) algorithms. The hybrid modified ABC-AFS algorithm was developed and integrated with artificial neural network (ANN) for landslide susceptibility mapping in Penang Island. It is noteworthy that this study represents the inaugural application of the hybrid modified ABC-AFS algorithm to a landslide susceptibility map prediction system. This novel approach integrates modified versions of the ABC and AFS algorithms with the ANN model, enabling the determination of optimal computational parameters for landslide susceptibility mapping. The performances of the proposed hybrid modified models are tested using optimization benchmark functions and it outperforms the individual modified ABC, modified AFS, ABC, and AFS models. Furthermore, the purposed models with ANN are assessed using receiver operating characteristic (ROC) curve and the obtained prediction accuracy value of the hybrid modified ABC-AFS-ANN model is 97.62%, which is the highest compared to modified ABC-ANN, modified AFS-ANN, ABC-ANN, AFS-ANN, and ANN models. Overall, all the modified models can produce satisfactory performance for landslide prediction of Penang Island, deducing t | |
| dc.identifier.uri | https://erepo.usm.my/handle/123456789/21546 | |
| dc.language.iso | en | |
| dc.title | Improved landslide susceptibility map prediction of penang island using hybrid artificial bee colony-artificial fish swarm models | |
| dc.type | Resource Types::text::thesis::doctoral thesis | |
| dspace.entity.type | Publication | |
| oairecerif.author.affiliation | Universiti Sains Malaysia |