input: , δ is a random, k points nearest to s are denoted as Nk(s), epochs is the number of training wheels, y is the actual true value and is an estimate value
output: The new dataset W after SMOTEENN processing
(1)
Generate new sample xnew and perform k_nearer neighbor calculation on xi;
(2)
Generate dataset W from a minority class samples of the input dataset Q;
(3)
for to range(m) do
(4)
Find the k minority class samples closer to xi;
(5)
W = [];
(6)
;
(7)
W.append (xnew);
(8)
return W;
(9)
Clean the newly generated dataset ;
(10)
Calculate the category t of s, according to the classification decision rules in Nk(s);