A Feature Weighted Fuzzy Clustering Algorithm Based on Multistrategy Grey Wolf Optimization

Algorithm 1

Semihomogenization and semirandomization.

	Input: population size n, dimension d of search space, interval [lj, uj] of variables Xj(j∈[l:d]) in search space.
	Output: initial population
(1)	for j = 1 to d
(2)	for i = 1 to n/2
(3)	randomly generate a value from [lj, uj] to assign to
(4)	end for
(5)	∆j=(uj-lj)/(n/2)
(6)	∇j = {[lj,lj+∆j], [lj+∆j,lj+2∆j], … [lj+(n/2–1)∆j,uj]}
(7)	for (n/2)+1 to n
(8)	randomly select a subinterval from the set ∇j, and randomly generate a value to assign to
(9)	update set ∇j: delete the subinterval selected in Step 8 from set ∇j
(10)	end for
(11)	end for
(12)	output the initial population {X^1,X^2,…X^n}