Advances in Artificial Neural Systems

Research Article

Hybrid Feature Selection Based Weighted Least Squares Twin Support Vector Machine Approach for Diagnosing Breast Cancer, Hepatitis, and Diabetes

Table 1

Algorithm for WLSTSVM classifier.
Linear WLSTSVMNonlinear WLSTSVM
Define and matrices as = [] and = [].
Calculate weight for each class using (1).
Choose penalty parameters and on the basis of validation.
Weight and bias required for the construction of nonparallel planes are calculated by using (4).
Generate two nonparallel planes by using (6).
For new data point, calculate its perpendicular distances from both the planes and a class is assigned to it by using (7). 		Define matrix as = .
Define kernel function and and matrices.
    = and
    = .  
Calculate weight for each class using (1).
Choose penalty parameters and on the basis of validation.
Weight and bias required for the construction of nonparallel planes are calculated by using (10) and (12).
Generate two nonparallel kernel generated surfaces by using (8).
For a new data point, calculate its perpendicular distances from both kernel generated surfaces and a class is assigned to it by using (14).