Journal article
Differential evolution memetic document clustering using chaotic logistic local search
Neural Information Processing, Vol.10634, pp.213-221
2017
Abstract
In this paper, we propose a Memetic-based clustering method that improves the partitioning of document clustering. Our proposed method is named as Differential Evolution Memetic Clustering (DEMC). Differential Evolution (DE) is used for the selection of the best set of cluster centres (centroids) while the Chaotic Logistic Search (CLS) is used to enhance the best set of solutions found by DE. For the purpose of comparison, the DEMC is compared with the basic DE, Differential Evolution Simulated Annealing (DESA) and the Differential Evolution K-Means (DEKM) methods as well as the traditional partitioning clustering using the K-means. The DEMC is also compared with the recently proposed Chaotic Gradient Artificial Bee Colony (CGABC) document clustering method. The reuters-21578, a pair of the 20-news group, classic 3 and TDT benchmark collection (TDT5) along with real-world six-event-crimes datasets are used in the experiments in this paper. The results showed that the proposed DEMC outperformed the other methods in terms of the convergence rate measured by the fitness function (ADDC) and the compactness of the resulted clusters measured by the F-macro and F-micro measures.
Details
- Title
- Differential evolution memetic document clustering using chaotic logistic local search
- Authors/Creators
- I. Al-Jadir (Author/Creator) - University of BaghdadK.W. Wong (Author/Creator) - Murdoch UniversityC.C. Fung (Author/Creator) - Murdoch UniversityH. Xie (Author/Creator) - Murdoch University
- Publication Details
- Neural Information Processing, Vol.10634, pp.213-221
- Publisher
- Springer Verlag
- Identifiers
- 991005542159507891
- Copyright
- © 2017 Springer International Publishing AG
- Murdoch Affiliation
- School of Engineering and Information Technology
- Language
- English
- Resource Type
- Journal article
Metrics
59 Record Views