A Quick Negative Selection Algorithm for One-Class Classification in Big Data Era

Algorithm 1

VorNSA .

Input: Training set , Self radius , Minimum detector radius

Output: Detector set

() normalize into

() construct voronoi diagram by sites

() get all cells in

() construct by

() foreach in

()   if has three or more same values in

()   then

() foreach in

()   compute the detector radius using Eq. (1)

()   if then

() foreach in

()   if has two same values in

()   then

() foreach in

()   compute the detector radius using Eq. (2)

()   if then

() return