A. Selamat et al. (Eds.): ACIIDS 2013, Part I, LNAI 7802, pp. 255–264, 2013.
© Springer-Verlag Berlin Heidelberg 2013
Performance of Different Techniques Applied in Genetic
Algorithm towards Benchmark Functions
Seng Poh Lim and Habibollah Haron
Department of Computer Science, Faculty of Computing,
Universiti Teknologi Malaysia,
81310 Skudai, Johor, Malaysia
lawrencess87@yahoo.com,
habib@utm.my
Abstract. Optimisation is the most interesting problems to be tested by using
Artificial Intelligence (AI) methods because different optimal results will be ob-
tained when different methods are implemented. Yet, there is no exact solution
from the methods implemented because random function is usually applied.
Genetic algorithm is a popular method which is used to solve the optimisation
problems. However, no any methods can execute perfectly because the way of
the method performs is different. Therefore, this paper proposed to compare the
performance of GA with different operation techniques by using the benchmark
functions. This can prove that different techniques applied in the operations can
let GA produces different result. Based on the experiment result, GA is proved
to perform well in the optimisation problems but it highly depends on the tech-
niques implemented. The techniques for each operation have shown different
performance in obtaining the time, minimum and average values for benchmark
functions.
Keywords: Genetic Algorithm, Optimisation, Benchmark Functions, Performance.
1 Introduction
Optimisation is the most interesting problems to be tested by using Artificial Intelli-
gence (AI) or Soft Computing methods. This is because different optimal results will
be obtained when different methods are implemented. However, there is no exact
solution from the methods implemented because random function is usually applied in
AI method. Basically, same optimisation problems were tested by using different
methods because the researchers want to test the performance of the method in obtain-
ing the result. If the result obtained is the most minimum and approximate to the ac-
tual optimum solution, hence this method is considered contains the best performance
compared to the others.
Genetic algorithm (GA) is a popular soft computing technique which is used to
solve the optimisation problems. Besides that, it can also be used to deal with schedul-
ing, surface fitting and searching problems. GA has been widely used to deal
with different kind of optimisation problems and it able to find and produce a good