Semisupervised SVM Based on Cuckoo Search Algorithm and Its Application

Algorithm 2

CCCS-S3VM.

(01) Input:

(02) Labeled dataset L

(03) Unlabeled dataset U

(04) Regularization parameter of unlabeled dataset

(05) Output:

(06) The label of unlabeled samples ;

(07) Final model CCCS-S3VM;

(08) Begin

(09) The regularization parameter and the kernel parameter of the labeled samples

   that obtain by CCCS-SVM algorithm;

 Training the initial model with labeled dataset L;

(11)   Predict the label of unlabeled dataset U using :

(12)  Initialization: ;

(13)  While   do

(14)   Solve formula (12) based on L, U, , , , obtain and ;

(15)   There is a pair of unlabeled samples and , Its label and is reverse,

    and the corresponding relaxation variables are satisfied: , it means that and

     likely to be a wrong label. The labels of the two are exchanged and the SVM

    problem is solved again. The approximate solution of the minimization of the

    objective function can be obtained after each round of iteration.

(16)  While do

(17)   ;    label exchange;

(18)    Solve formula (12) based on L, U, , obtain and ;

(19)  End while

(20)  

(21)  End while

(22) End