Automatic clustering based on nature inspired metaheuristics

• Grant number: 17/06142-2
• Support Opportunities: Scholarships in Brazil - Scientific Initiation
• Effective date (Start): June 01, 2017
• Effective date (End): May 31, 2018
• Field of knowledge: Physical Sciences and Mathematics - Computer Science - Computing Methodologies and Techniques
• Principal Investigator: Adriane Beatriz de Souza Serapião
• Grantee: Maynara Natalia Scoparo
• Host Institution: Instituto de Geociências e Ciências Exatas (IGCE). Universidade Estadual Paulista (UNESP). Campus de Rio Claro. Rio Claro , SP, Brazil

Abstract

Data clustering is one of the most important unsupervised techniques of data management, which is used in many scientific and engineering applications, such as machine learning, data mining, pattern recongnitionand image processing. It consists of splitting a dataset into smaller subsets, named clusters. The partition of datasets is obtained by establishing a function that assigns each object of the dataset to subset, so that similar objects are in the same cluster. A fundamental challenge in clustering analysis is to determine the best estimate of the number of clusters, which is recognized as the automatic clustering problem. The difficulty of choosing the appropriate number of clusters is due to the lack of previous knowledge about the application's domain, especially when the data have many dimensions, when the clusters differ widely distinct in shape, size and density and when there is overlap between groups. In this project, three Swarm Intelligence algorithms willed used for the automatic clustering problem in numeric datasets. Such algorithms will be developed to optimize division criteria, using clustering measures, in order to find the optimal number of clusters and the centroids coordinates. The bioinspired optimization methods Whale Optimization Algorithm, Cuckoo Search and Cat Swarm Optimization will be adapted for the clustering task by using the partitioned approach. In order to evaluate the results of these algorithms for automatic clustering, internal and external validation indexes will be used.