Competitive Particle Swarm Optimization for Multi-Category Text Feature Selection

. 2019 Jun 18;21(6):602. doi: 10.3390/e21060602

Algorithm 1 Competitive particle swarm optimization.

1:
input: $m_{c}, m_{f}, v$ ; ▹ The number of particles for each group $m_{c}, m_{f}$ , the maximum number of FFCs v
2:
output:S; ▹ the final feature subset S
3:
$[C, F] \leftarrow$ initialization $(m_{c}, m_{f})$ ; ▹ initialize particles using Algorithm 2
4:
$[S_{c}, S_{f}] \leftarrow$ generate subsets based on $C, F$ ; ▹ use locations of particles
5:
$[E_{c}, E_{f}] \leftarrow$ evaluate subsets $S_{c}, S_{f}$ ; ▹ evaluate subsets using fitness function
6:
$u \leftarrow m_{c} + m_{f}$ ; ▹ set u to the number of whole particles
7:
while $u < v$ do
8:
update C using Equations (2) and (3); ▹ update locations of particles
9:
$[S_{c}, S_{f}] \leftarrow$ generate subsets based on $C, F$ ;
10:
$[E_{c}, E_{f}] \leftarrow$ evaluate subsets $S_{c}, S_{f}$ ;
11:
$u \leftarrow u + m_{c} + m_{f}$ ;
12:
$[C, F, m_{c}, m_{f}] \leftarrow$ competition $(C, F, E_{c}, E_{f}, m_{c}, m_{f})$ ; ▹ use Algorithm 3
13:
$S \leftarrow$ the best feature subset so far;
14:
end while