The Optimisation for Local Coupled Extreme Learning Machine Using Differential Evolution

Algorithm 2

Evolutionary local coupled extreme learning machine.

Require:

   , the training set ;

   , the activation function;

   , the similarity function;

   , the fuzzy membership function;

   , the number of hidden nodes;

   Itermax, the preset maximum learning epoch of DE.

(1) Randomly designate hidden node parameters (, ), hidden node address , and radius .

(2) .

(3) while     do

(4) (1)  Calculate the hidden layer output matrix ,

   (2)  Calculate the output weight ,

   (3)  Adjust (, ) using DE,

   (4)  .

(5) end while

(6) (1)  Calculate the hidden layer output matrix .

   (2)  Calculate the output weight .