() Input , , , , samples, number of subsets , the number of clustering centers |
and population size ; |
() Initialize population , generation number ; |
() Calculate the fitness of each individual in the initial population according to Eq. (8); |
() For all to do |
() Select survivor according to the wheel selection rule [23]; |
() End for |
() For all to do |
() For all to do |
() Select randomly two individuals and calculate the crossover probability according to Eq. (12); |
() Generate a random floating-point number , ; |
() If then |
() Perform arithmetic crossover operation [24]; |
() End if |
() End for |
() For all to do |
() Calculate the mutation probability according to Eq. (14); |
() Generate a random floating-point number , ; |
() If then |
() Perform Gaussian mutation operation [25]; |
() End if |
() End for |
() Calculate the fitness of each individual in the new population ; |
() If then |
() Return individuals with the highest fitness; |
() Else if |
() Return to selecting survivor; |
() End if |
() End for |
() Output final population |
() Initialize obtained the excellent clustering centers, the number of iterations , the error log , |
and the cut-off error ; |
() Establish new objective function according to Eq. (16), (17) and (18); |
() For to |
() If |
() Establish new membership degree matrix according to Eq. (21); |
() Obtain new formula of clustering centers according to Eq. (22); |
() , ; |
() Else if |
() Convergence; |
() End if |
() End for |