Hybrid ant bee colony algorithm (HABC) for classifiation and prediction mission
A combine social insect's movement attracts scientists toward new solutions for different mathematical and statistical problems. Chief among of them are the Artificial Bee Colony (ABC) algorithm and Ant Colony Optimization (ACO) algorithm that simulate the intelligent foraging behaviours of hon...
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Conference or Workshop Item |
| Published: |
2011
|
| Subjects: | |
| Online Access: | http://eprints.uthm.edu.my/2963/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | A combine social insect's movement attracts scientists toward new solutions for different mathematical and statistical problems. Chief among of them are the Artificial Bee Colony (ABC) algorithm and Ant Colony Optimization (ACO) algorithm that simulate the intelligent foraging behaviours of honey bee and ant swarms. These algorithms have been successfully used in many different tasks such as classification, global optimization for numerical function, image segmentation and optimization of Artificial Neural Algorithm (ANNs) weights. Multilayer perception (MLP), the widely known ANNs was normally trained with the standard Back-Propagation (BP) algorithm for minimizing the network error. However, using the BP algorithm for training the MLP always contribute to a problem of suboptimal weights because of the proposed, and it is called Hybrid Bee Ant Colony (BBAC) algorithm. In this work, HBAC is used for training the MLP and XOR classification tasks. The performance of the proposed HBAC algorithm is benchmarked against the standard BP. Experimental results show that the proposed HBAC algorithm outperformed the BP and ACO algorithms when used to train the MLP with lower prediction error. |
|---|