Table 1

Algorithm: Pseudo-code of QOCNNA.

Algorithm: Pseudocode of QOCNNA
Initialize the initial parameters of the QOCNNA (Nr, Tmax, Jr, and K)
% QOBL-based population initialization %
Randomly generate an initial population X
Generate quasi-opposite solutions QOX of initial pattern solutions X
Calculate the fitness function values of the combined set {X,QOX} and sort them
Select the Nr ​ best solutions from the set {X,QOX} as the initial population
Randomly generate the weight matrix considering the imposed constraints in equation (39)
Define target solution Xtarget and its corresponding target weight (Wtarget )
for t=1: Tmax
  % Search process of NNA %
  Generate new pattern solutions (Xnew) using equations (41) and (42)
  Update the weight matrix (W) using equation (43)
  for i=1: Nr
   if rand<β
    Perform the bias operator
   else (rand≥β)
    Apply the transfer function operator
   end if
  end for
  Calculate the fitness function values for all updated pattern solutions
  Update the value of β\betaβ
  % QOBL-based generation jumping %
  if rand< Jr
   Generate quasi-opposite solutions (QOX) of updated pattern solutions Xt
   Calculate the fitness function values for quasi-opposite solutions (QOX)
   if fi(QOXt )<fi(Xt ​)
    Xt = QOXt
    fi(QOXt )<fi(Xt )
   end if
  end if
  % Chaotic local search %
  Update the target solution and its corresponding target weight
  Perform the CLS strategy to generate a better target solution
end for
Return the target solution Xtarget ​ and its corresponding fitness function value

Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.

Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.

Initial download of the metrics may take a while.