An Al-Biruni Earth Radius Optimization-Based Deep Convolutional Neural Network for Classifying Monkeypox Disease

View full-text article in PMC

. 2022 Nov 21;12(11):2892. doi: 10.3390/diagnostics12112892

Algorithm 3 Proposed BERSFS algorithm

1:
Initialize BERSFS population $S_{i} (i = 1, 2, \dots, d)$ with size d, iterations $T_{m a x}$ , fitness function $F_{n}$ , $t = 1$ , BERSFS parameters
2:
Calculate fitness function $F_{n}$ for each $S_{i}$
3:
Find best solution as $S^{*}$
4:
while $t \leq T_{m a x}$ do
5:
if ( ${r a n d}_{B E R S F S} > 0.5$ ) then
6:
for ( $i = 1 : i < n_{1} + 1$ ) do
7:
Update $r_{1} = h_{1} \frac{c o s (x)}{1 - c o s (x)}$ , $r_{2} = h_{2} \frac{c o s (x)}{1 - c o s (x)}$
8:
Calculate $D = r_{1} (S (t) - 1)$
9:
Update positions to head toward best solution as

$S (t + 1) = S (t) + D (2 r_{2} - 1)$
10:
end for
11:
for ( $i = 1 : i < n_{2} + 1$ ) do
12:
Update $r = h \frac{c o s (x)}{1 - c o s (x)}$ , $r_{3} = h_{3} \frac{c o s (x)}{1 - c o s (x)}$
13:
Calculate $D = r_{3} (L (t) - S (t))$
14:
Update positions Elitism of best solution as

$S (t + 1) = r^{2} (S (t) + D)$
15:
Calculate $k = 1 + \frac{2 \times t^{2}}{M a x_{i t e r}^{2}}$
16:
Update positions Investigating area around best solution as

$S^{'} (t + 1) = r (S^{*} (t) + k)$
17:
Compare $S (t + 1)$ and $S^{'} (t + 1)$ to select best solution $S^{*}$
18:
if best fitness is not changed for last two iterations then
19:
Mutate solution as $S (t + 1) = k * z^{2} - h \frac{c o s (x)}{1 - c o s (x)}$
20:
end if
21:
end for
22:
else
23:
for ( $i = 1 : i < n + 1$ ) do
24:
Calculate updated best solution as

$S^{'}^{*} (t + 1) = G a u s s i a n (μ_{S^{*}}, σ) + (η \times S^{*} (t) - η^{'} \times P_{i})$
25:
end for
26:
end if
27:
Update the fitness function $F_{n}$ for each $S_{i}$
28:
Find best solution as $S^{*}$
29:
Update BERSFS parameters, $t = t + 1$
30:
end while
31:
Return $S^{*}$