Skip to main content

Table 1 The best parameters setting given the best cost solution for all proposed algorithms

From: Integration of evolutionary computation algorithms and new AUTO-TLBO technique in the speaker clustering stage for speaker diarization of broadcast news

GA parameters PSO parameters
Maximum number of iterations = 200
 Population size(nPop) = 100 Constriction coefficients
 Crossover percentage = 0.7 phi1 = 2.05, phi2 = 2.05
phi = phi1 + phi2
 Number of offsprings (nc) = 2*round(pc*nPop/2)
 Mutation percentage (pm) = 0.3 chi = 2/(phi-2 + sqrt(phi^2-4*phi))
 Number of mutants (nm) = round (pm*nPop) Inertia weight w = chi
 Mutation rate (mu) = 0.02 Inertia weight damping ratio (wdamp) = 1
 Selection pressure (beta) = 8 Personal learning coefficient (c1) = chi*phi1
 Gamma = 0.2 Global learning coefficient (c2) = chi*phi2
  Velocity maximal = 0.1*(VarMax-VarMin)
  Velocity minimal = − VelMax
  VarMin = − 10; VarMax = 10
DE parameters
 Maximum number of iterations (MaxIt) = 200
 Population size (nPop) = 50
 Lower bound of scaling factor (beta_min) C r min = 0.2
 Upper bound of scaling factor (beta_max)  C r max = 0.8
 Crossover probability (pCR) = 0.2
TLBO parameters
 MaxIt = 1000; nPop = 50; T F  = 1