Cancer Feature Selection and Classification Using a Binary Quantum-Behaved Particle Swarm Optimization and Support Vector Machine

Pseudocode 3

The psudocode the BPSO/SVM.

Processing of data set;

Initialize the current positions and the pbest positions of all particles which are binary bits with each representing whether the

corresponding gene is selected or not;

do

Determine the mean best position among the particles by mbest = Get_mbest(pbest), select a suitable value for ;

 for   to population size

   Call the LIBSVM tool box to construct the SVM classifier and get the classification accuracy for the data;

  With the classification accuracy and the number of selected genes (i.e. the number of features given by the number of bits

  with value 1), evaluate the objective function value according to Section 3.3;

   Update , and , it means

    if   then ;

  and , ;

  then get a stochastic position by = Get_P (, best)

    for   to dimensionality

    Compute the mutation probability ;

    Generate the new substring by = Transf(, );

   and get the new position by combining all new substring ()

  endfor

 endfor

until termination criterion is met;

Output the best solution which have been found (best)