DECSAI A Greedy Randomized Adaptive Search Procedure for the Clustering Problem
نویسندگان
چکیده
The aim of this paper is to present a new proposal for Cluster Analysis based on a Greedy Randomized Adaptive Search Procedure (GRASP), with the objective of overcoming the convergence to a local solution. It uses a probabilistic greedy Kaufman initialization method for getting initial solutions and the K-Means algorithm as a local search algorithm. The new proposal will become a new initialization approach to K-Means. Hence, we have compared some initialization methods for the K-Means algorithm: Random, Forgy, Macqueen and Kaufman, with the GRASP. Our results suggest that the Kaufman initialization method performs better than the other three methods and the hybrid GRASP – K-Means with Kaufman inicialization improves these results even more. The proposed method obtains high quality solutions for the benchmark problems considered.
منابع مشابه
Using Greedy Randomize Adaptive Search Procedure for solve the Quadratic Assignment Problem
Greedy randomize adaptive search procedure is one of the repetitive meta-heuristic to solve combinatorial problem. In this procedure, each repetition includes two, construction and local search phase. A high quality feasible primitive answer is made in construction phase and is improved in the second phase with local search. The best answer result of iterations, declare as output. In this stu...
متن کاملFitting the Three-parameter Weibull Distribution by using Greedy Randomized Adaptive Search Procedure
The Weibull distribution is widely employed in several areas of engineering because it is an extremely flexible distribution with different shapes. Moreover, it can include characteristics of several other distributions. However, successful usage of Weibull distribution depends on estimation accuracy for three parameters of scale, shape and location. This issue shifts the attentions to the requ...
متن کاملA hybrid metaheuristic using fuzzy greedy search operator for combinatorial optimization with specific reference to the travelling salesman problem
We describe a hybrid meta-heuristic algorithm for combinatorial optimization problems with a specific reference to the travelling salesman problem (TSP). The method is a combination of a genetic algorithm (GA) and greedy randomized adaptive search procedure (GRASP). A new adaptive fuzzy a greedy search operator is developed for this hybrid method. Computational experiments using a wide range of...
متن کاملA Hybrid Clustering Algorithm Based on Honey Bees Mating Optimization and Greedy Randomized Adaptive Search Procedure
This paper introduces a new hybrid algorithmic nature inspired approach based on the concepts of the Honey Bees Mating Optimization Algorithm (HBMO) and of the Greedy Randomized Adaptive Search Procedure (GRASP), for optimally clustering N objects into K clusters. The proposed algorithm for the Clustering Analysis, the Hybrid HBMO-GRASP, is a two phase algorithm which combines a HBMO algorithm ...
متن کاملMulti - Start Methods for the Capacitated Clustering Problem
In this work, we investigate the adaptation of the Greedy Randomized Adaptive Search Procedure (GRASP) and Iterated Greedy methodologies to the Capacitated Clustering Problem (CCP). In particular, we focus on the effect of the balance between randomization and greediness on the performance of these multi-start heuristic search methods when solving this NP-hard problem. The former is a memory-le...
متن کامل