A Novel Heuristic Approach for the Simultaneous Selection of the Optimal Clustering Method and Its Internal Parameters for Time Series Data
No Thumbnail Available
Identifiers
Publication date
2020
Advisors
Journal Title
Journal ISSN
Volume Title
Publisher
Springer Verlag
Citations
Export
Abstract
Clustering methods have become popular in the last years due to the need of analyzing the high amount of collected data from different fields of knowledge. Nevertheless, the main drawback of clustering is the selection of the optimal method along with its internal parameters in an unsupervised environment. In the present paper, a novel heuristic approach based on the Harmony Search algorithm aided with a local search procedure is presented for simultaneously optimizing the best clustering algorithm (K-means, DBSCAN and Hierarchical clustering) and its optimal internal parameters based on the Silhouette index. Extensive simulation results show that the presented approach outperforms the standard clustering configurations and also other works in the literature in different Time Series and synthetic databases.
Description
Publisher Copyright: © 2020, Springer Nature Switzerland AG.
Type
Citation
Navajas-Guerrero , A , Manjarres , D , Portillo , E & Landa-Torres , I 2020 , A Novel Heuristic Approach for the Simultaneous Selection of the Optimal Clustering Method and Its Internal Parameters for Time Series Data . in F Martínez Álvarez , A Troncoso Lora , J A Sáez Muñoz , E Corchado & H Quintián (eds) , 14th International Conference on Soft Computing Models in Industrial and Environmental Applications SOCO 2019, Proceedings . Advances in Intelligent Systems and Computing , vol. 950 , Springer Verlag , pp. 179-189 , 14th International Conference on Soft Computing Models in Industrial and Environmental Applications, SOCO 2019 , Seville , Spain , 13/05/19 . https://doi.org/10.1007/978-3-030-20055-8_17
conference
conference