A Novel Data Classifier Using Social Spider Optimization
Ravichandran Thalamala1, B. Janet2, A.V. Reddy3
1Ravichandran Thalamala, Department of Computer Applications, National Institute of Technology, Trichy-620015, India.
2B. Janet, Department of Computer Applications, National Institute of Technology, Trichy-620015, India.
3A.V. Reddy, Department of Computer Applications, National Institute of Technology, Trichy-620015, India.
Manuscript received on 02 June 2019 | Revised Manuscript received on 10 June 2019 | Manuscript published on 30 June 2019 | PP: 1756-1773 | Volume-8 Issue-8, June 2019 | Retrieval Number: H6600068819 /19©BEIESP
Open Access | Ethics and Policies | Cite | Mendeley | Indexing and Abstracting
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC-BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: In current decade, Social Spider Optimization (SSO) has become popular among researchers due to its abilityto represent and handle very high and complex dimensional solution space. Like the other nature inspired algorithms,it also takes inspiration from nature. It mimics the cooperative behavior of social spiders in the forests. Unlike theother nature inspired algorithms, its agents have gender property due to which the algorithm maintains the balancebetween exploration and exploitation. Recently, a few researchers have employed SSO for clustering data. In this article,we propose a new classification algorithm called All Prototypes Social Spider Optimization for Data Classification(APSSODC) in which each spider has the prototypes of all data instances of the dataset. As the dimensionality ofsolution space in APSSODC is very high and equal to the product of degree and cardinality of the dataset, we proposeanother algorithm called Single Prototype Social Spider Optimization for Data Classification (SPSSODC) that reducesthe dimensionality of the solution space. It considers each spider as a single prototype of a data instance present in thedataset. Wefound that SPSSODC outperforms the existing algorithms including APSSODC with respect to classificationaccuracy.
Key words: Nature Inspired Algorithms, Data Classification, Social Spider Optimization, Solution Space, Lassificationaccuracy
Scope of the Article: Classification