GRASP with path-relinking for data clustering: a case study for biological data

Authors:
Rafael M. D. Frinhani;Ricardo M. A. Silva;Geraldo R. Mateus;Paola Festa;Mauricio G. C. Resende
Affiliations:
Universidade Federal de Minas Gerais, Belo Horizonte, MG, Brazil;Universidade Federal de Lavras, Lavras, MG, Brazil and Universidade Federal de Pernambuco, Recife, PE, Brazil;Universidade Federal de Minas Gerais, Belo Horizonte, MG, Brazil;University of Napoli Federico II, Napoli, Italy;AT&T Labs Research, Florham Park, NJ
Venue:
SEA'11 Proceedings of the 10th international conference on Experimental algorithms
Year:
2011

Citing 7
Cited 0

Mersenne twister: a 623-dimensionally equidistributed uniform pseudo-random number generator

ACM Transactions on Modeling and Computer Simulation (TOMACS) - Special issue on uniform random number generation
Tabu Search

Tabu Search
Grasp and Path Relinking for 2-Layer Straight Line Crossing Minimization

INFORMS Journal on Computing
A Hybrid GRASP with Perturbations for the Steiner Problem in Graphs

INFORMS Journal on Computing
Parallel GRASP with path-relinking for job shop scheduling

Parallel Computing - Special issue: Parallel computing in numerical optimization
GRASP with Path Relinking for Three-Index Assignment

INFORMS Journal on Computing
Investigation of a new GRASP-based clustering algorithm applied to biological data

Computers and Operations Research

Quantified Score

Hi-index	0.00

Visualization

Abstract

Cluster analysis has been applied to several domains with numerous applications. In this paper, we propose several GRASP with path-relinking heuristics for data clustering problems using as case study biological datasets. All these variants are based on the construction and local search procedures introduced by Nascimento et. al [22]. We hybridized the GRASP proposed by Nascimento et. al [22] with four alternatives for relinking method: forward, backward, mixed, and randomized. To our knowledge, GRASP with path-relinking has never been applied to cluster biological datasets. Extensive comparative experiments with other algorithms on a large set of test instances, according to different distance metrics (Euclidean, city block, cosine, and Pearson), show that the best of the proposed variants is both effective and efficient.