Name: Mario Romulo de Brito Fernandes
Type: MSc dissertation
Publication date: 20/09/2017
Advisor:

Namesort descending Role
Thomas Walter Rauber Advisor *

Examining board:

Namesort descending Role
Elias Silva de Oliveira Internal Examiner *
Thomas Walter Rauber Advisor *

Summary: This essay introduces a novel hybrid coevolutionary algorithm, called Zombie Search
(ZS) to locate the global minimum for unconstrained optimization problems. ZS works
with two populations simulating a coevolutionary and antagonistic predator-prey process
between humans and zombies. This process combines the benefits of swarm intelligence,
success history based parameter adaptation and estimation of distribution algorithms. The
algorithm applies three types of different movements for each population, using probabilistic
models based on Gaussian and Cauchy distributions, aiming at obtaining higher quality
solutions. In order to evaluate the performance of ZS, experiments were performed for
a set of public benchmark functions (CEC’15). To promote better performance, the ZS
parameters are self-optimized using an uncertainty handling method for each benchmark
function present in CEC’15. The results obtained by ZS are compared to other six stateof-the-art algorithms. The results suggest that the proposal is competitive and partially
returns better results when the problem has a small number of variables, according to the
analyzes used.

Transparência Pública
Acesso à informação

© 2013 Universidade Federal do Espírito Santo. Todos os direitos reservados.
Av. Fernando Ferrari, 514 - Goiabeiras, Vitória - ES | CEP 29075-910