2015 International Symposium of Computer Application and Information Technology (IsCait 2015)

242 / 1971-01-01 00:00:00

Cancer Classification And Biomaker Discovery Based Hybrid Bqpso/svm Algorithm

BQPSO,BPSO,GA,SVM,LOOCV

In this work, a BQPSO/SVM (Quantum-Behaved Particle Swarm Optimization with Binary Encoding) algorithm for cancer feature selection is proposed. We also implement BPSO/SVM (Particle Swarm Optimization) and GA/SVM (Genetic Algorithm) to be compared with the proposed algorithm. All these three are augmented with Support Vector Machines (SVM) with Leave-one-out Cross Validation (LOOCV) and assessed on five microarray data sets (Leukemia, Prostate, Colon, Lung, Lymphoma). The results show that BQPSO/SVM has significant advantages in accuracy, robustness and the number of feature genes selected compared with the other two algorithms.