TY - JOUR
T1 - A Novel Artificial Bee Colony Optimization Algorithm with SVM for Bio-inspired Software-Defined Networking
AU - Chiang, Hsiu Sen
AU - Sangaiah, Arun Kumar
AU - Chen, Mu Yen
AU - Liu, Jia Yu
N1 - Funding Information:
The authors would like to thank the Ministry of Science and Technology of Taiwan for Grants MOST106-2410-H-025-007, MOST-103-2410-H-025-022-MY2 and National Science Council of Taiwan for Grants NSC-101-2410-H-025-004-MY2 which supported part of this research.
PY - 2020/4/1
Y1 - 2020/4/1
N2 - In recent years, artificial intelligence and bio-inspired computing methodologies have risen rapidly and have been successfully applied to many fields. Bio-inspired network systems are a field of biology and computer science, it has the high relation to the bio-inspired computing and bio-inspired system. It has the self-organizing and self-healing characteristics that help them in achieving complex tasks with much ease in the network environment. Software-defined networking provides a breakthrough in network transformation. However, increasing network requirement and focus on the controller for determining the network functionality and resources allocations aims at self-management capabilities. More recently, the artificial bee colony (ABC) algorithm has been used to solve the issues of parameter optimization. In this paper, a discretized food source for an artificial bee colony (DfABC) optimization algorithm is proposed and applied to optimize the kernel parameters of a support vector machine (SVM) model, creating a new hybrid. In order to further improve prediction accuracy, the proposed DfABC algorithm is applied to six popular UCI datasets. We also compare the DfABC algorithm to particle swarm optimization (PSO), the genetic algorithm (GA), and the original ABC algorithm. The experimental results show that the proposed DfABC-SVM model achieves better classification accuracy with a shorter convergence time, outperforming the other hybrid artificial intelligence models.
AB - In recent years, artificial intelligence and bio-inspired computing methodologies have risen rapidly and have been successfully applied to many fields. Bio-inspired network systems are a field of biology and computer science, it has the high relation to the bio-inspired computing and bio-inspired system. It has the self-organizing and self-healing characteristics that help them in achieving complex tasks with much ease in the network environment. Software-defined networking provides a breakthrough in network transformation. However, increasing network requirement and focus on the controller for determining the network functionality and resources allocations aims at self-management capabilities. More recently, the artificial bee colony (ABC) algorithm has been used to solve the issues of parameter optimization. In this paper, a discretized food source for an artificial bee colony (DfABC) optimization algorithm is proposed and applied to optimize the kernel parameters of a support vector machine (SVM) model, creating a new hybrid. In order to further improve prediction accuracy, the proposed DfABC algorithm is applied to six popular UCI datasets. We also compare the DfABC algorithm to particle swarm optimization (PSO), the genetic algorithm (GA), and the original ABC algorithm. The experimental results show that the proposed DfABC-SVM model achieves better classification accuracy with a shorter convergence time, outperforming the other hybrid artificial intelligence models.
UR - http://www.scopus.com/inward/record.url?scp=85052138216&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85052138216&partnerID=8YFLogxK
U2 - 10.1007/s10766-018-0594-6
DO - 10.1007/s10766-018-0594-6
M3 - Article
AN - SCOPUS:85052138216
SN - 0885-7458
VL - 48
SP - 310
EP - 328
JO - International Journal of Parallel Programming
JF - International Journal of Parallel Programming
IS - 2
ER -