“…On the basis of SVM and GEPSVM, Jayadeva et al proposed a novel binary classifier, Twin Support Vector Machine (TWSVM), which classifies the patterns of two classes by using two non-parallel hyper-planes [22]. TWSVM solves a pair of QPPs instead of single complex QPP as in SVM which makes the learning of TWSVM four times faster as compared to conventional SVM [22][23]. In SVM, all patterns together provide constraints to QPP, while in TWSVM patterns of one of the two classes provide constraints to each QPP.…”