Designing Artificial Neural Networks Using Particle Swarm Optimization Algorithms

Algorithm 1

New Model of PSO pseudocode.

 ()  Given a population of , individuals.

 ()  Initialize the population at random.

 ()  Until a stop criteria is reached:

 ()  if   is not improved during   then

 ()    Perform a velocity resting, (7).

 ()  end if

 ()  if    then

 ()    Create new neighbourhoods.

 ()  end if

() Modify inertia weight, (6).

() for each individual    do

()  Evaluates their fitness.

() end for

() for each individual   do

()  Update its best position  .

() end for

() for each neighbourhood   do

()  Update the best individual  .

() end for

() for each individual and each dimension: do

()  Compute the velocity update equation  , (11).

()  Compute the current position  .

()  if    then

()   Apply crossover operator, (8) and (9).

()  end if

()  if    then

()   Apply mutation operator, (10).

()   end if

() end for