Local Similarity-Based Fuzzy Multiple Kernel One-Class Support Vector Machine

Algorithm 1

Local similarity-based memberships.

	Input: the training set , the kernel function set ,
	The kernel weight ,
	Output: the membership vector
(1)	Preprocess the training set
(2)	Calculate the combined kernel matrix according to equation (10)
(3)	Sort from large to small as ,
(4)	Calculate the constant according to , and fixed threshold according to
(5)	for i = 1 : l do
(6)	initialize
(7)	for j = 1 : l do
(8)	if then
(9)
(10)	end if
(11)	end for
(12)	end for
(13)	Calculate
(14)	for i = 1 : l do
(15)	Calculate the membership degree of the sample according to equation (22)
(16)	end for
(17)	end