Hierarchical Clustering Approach With Hybrid Genetic Algorithm for Combinatorial Optimization Problems

Abstract

Engineering field has inherently many combinatorial optimization problems which are hard to solve in some definite interval of time especially when input size is big. Although traditional algorithms yield most optimal answers, they need large amount of time to solve the problems. A new branch of algorithms known as evolutionary algorithms solve these problems in less time. Such algorithms have landed themselves for solving combinatorial optimization problems independently, but alone they have not proved efficient. However, these algorithms can be joined with each other and new hybrid algorithms can be designed and further analyzed. In this paper, hierarchical clustering technique is merged with IAMB-GA with Catfish-PSO algorithm, which is a hybrid genetic algorithm. Clustering is done for reducing problem into sub problems and effectively solving it. Results taken with different cluster sizes and compared with hybrid algorithm clearly show that hierarchical clustering with hybrid GA is more effective in obtaining optimal answers than hybrid GA alone.

Materialized View Selection Based on Adaptive Genetic Algorithm and Its Implementation with Apache Hive

“…A genetic algorithm (GA) is a search heuristic that mimics the process of natural selection [8]. The parallelism, self-study nature and robustness of the genetic algorithm make it very effective in solving the combinatorial optimization problem [9]. We apply an adaptive genetic algorithm to solve the materialized view selection problem.…”