%0 Journal Article
	%A Hassan Zarei and  Ali Vahidian Kamyad and  Sohrab Effati
	%D 2011
	%J International Journal of Medical and Health Sciences
	%B World Academy of Science, Engineering and Technology
	%I Open Science Index 55, 2011
	%T An Adaptive Memetic Algorithm With Dynamic Population Management for Designing HIV Multidrug Therapies
	%U https://publications.waset.org/pdf/4022
	%V 55
	%X In this paper, a mathematical model of human immunodeficiency
virus (HIV) is utilized and an optimization problem is
proposed, with the final goal of implementing an optimal 900-day
structured treatment interruption (STI) protocol. Two type of commonly
used drugs in highly active antiretroviral therapy (HAART),
reverse transcriptase inhibitors (RTI) and protease inhibitors (PI), are
considered. In order to solving the proposed optimization problem an
adaptive memetic algorithm with population management (AMAPM)
is proposed. The AMAPM uses a distance measure to control the
diversity of population in genotype space and thus preventing the
stagnation and premature convergence. Moreover, the AMAPM uses
diversity parameter in phenotype space to dynamically set the population
size and the number of crossovers during the search process.
Three crossover operators diversify the population, simultaneously.
The progresses of crossover operators are utilized to set the number
of each crossover per generation. In order to escaping the local optima
and introducing the new search directions toward the global optima,
two local searchers assist the evolutionary process. In contrast to
traditional memetic algorithms, the activation of these local searchers
is not random and depends on both the diversity parameters in
genotype space and phenotype space. The capability of AMAPM in
finding optimal solutions compared with three popular metaheurestics
is introduced.
	%P 255 - 262