Page 193 - Handbook of Deep Learning in Biomedical Engineering Techniques and Applications

182 Chapter 6 Plant leaf disease classification based on feature selection

Initialize the particle population
Initialize parameters
while (t< Max number of iteration)

for each particle with position x p
calculate fitness value f(x p)
if f(x p) is better than pbest p then
pbest p x p
endif
if f(pbest p) is better than gbest then
gbest pbest p
endif
end for
update w according to equation (16)
for each particle with position x p
update c1, c2 according to equation (14), (15)
calculate velocity of each particle by equation (17)
update position of each particle by equation (18)
end for
if rand (0,1) < prob
run GWO
x p = position of the best wolf
endif
t=t+1
end while
return gbest
Figure 6.23 Pseudocode for APGWO.

4.3.2.4 Wrapper-based adaptive particleegray wolf optimization
The solution for the wrapper is a binary array, with dimension
of 1 n, where n is the total number of features. Selected features
will take value of 1 and 0 otherwise (Fig. 6.24).
The parameters set for different algorithms are as follows: 20
search agents (for PSO main loop), 20 search agents (for nested
GWO loop), 20 iterations for main PSO loop, 5 iterations for
nested GWO, wMax ¼ 0.9, wMin ¼ 0.2.