Fish School Search Algorithm on GPU to the task of data clustering analysis

• Grant number: 13/08741-0
• Support Opportunities: Scholarships in Brazil - Scientific Initiation
• Effective date (Start): June 01, 2013
• Effective date (End): May 31, 2014
• Field of knowledge: Physical Sciences and Mathematics - Computer Science - Computing Methodologies and Techniques
• Principal Investigator: Adriane Beatriz de Souza Serapião
• Grantee: Guilherme Sanchez Corrêa
• Host Institution: Instituto de Geociências e Ciências Exatas (IGCE). Universidade Estadual Paulista (UNESP). Campus de Rio Claro. Rio Claro , SP, Brazil

Abstract

Cluster analysis is a collection of unsupervised computational techniques often used in applications related to patterns search, such as Data Mining. Clustering algorithms aim to separate objects into useful or significative groups (clusters), according to the features of the object, so as to maximize the similarity of objects within a group and minimize the similarity between objects of different groups, using pre-defined metrics. This project aims to adapt recent bioinspired optimization methods, called Fish School Search, to perform the numerical data clustering task. The proposal is to use this method along with the bioinspired approach for partitioning groups. Two different implementations to examine the clustering algorithm are proposed: one for sequential and one parallel CPU with GPU programming in a CUDA environment, the results will be compared with each other to assess performance in the task of grouping data. (AU)

Scientific publications

(References retrieved automatically from Web of Science and SciELO through information on FAPESP grants and their corresponding numbers as mentioned in the publications by the authors)

SERAPIAO, ADRIANE B. S.; CORREA, GUILHERME S.; GONCALVES, FELIPE B.; CARVALHO, VERONICA O.. Combining K-Means and K-Harmonic with Fish School Search Algorithm for data clustering task on graphics processing units. APPLIED SOFT COMPUTING, v. 41, p. 290-304, APR 2016. (13/23027-1, 13/08730-8, 13/08741-0)