Page 189 - Handbook of Deep Learning in Biomedical Engineering Techniques and Applications

178 Chapter 6 Plant leaf disease classification based on feature selection

its herb’s flying experience. Each individual is considered as a
particle or a point in the D-dimensional search space. The i th
2
k

1
particle is represented as X i ¼ x ; x ; .; x . The best previous
i
i
i
position that gives the best fitness value of the i th particle is
recorded and represented as in p best . The index of the best particle
i
in the population is noted as g k . The velocity that defines the
best
rate of the position changing for particle i is represented as
1 2 k
V i ¼ v ; v ; .; v (Fig. 6.21).
i i i
The mathematical model of PSO is represented as the
following equations:

t
t
v tþ1 ¼ w v þ c 1 rand pbest x t þ c 2 rand
k k k k
t
gBest x (6.3)
k
t
x tþ1 ¼ x þ v t (6.4)
k k k
where c 1 and c 2 are two constants, and rand is any random num-

t
ber between [0, 1]. The term c 1 rand pbest x t is called
k k
t
“cognitive component,” whereas c 2 rand gBest x is
k
referred as “social component.”

Figure 6.21 Position updating in PSO. PSO, particle swarm optimization.