Particle Swarm Optimization Feature Selection for Classification of Survival Analysis in Cancer
It could be seen that almost survival analysis in biomedical and healthcare are focusing on survival time; related to how long individuals with disease will survive or dying. While most existing survival analysis aims at improving the survival rate by extracting the useful knowledge from patient data, either through existing techniques or through development of new techniques, this paper focused on selecting only the significant and relevant patient data with minimal information loss to classify the patient survival. This paper highlights and discusses the concept and limitation of classification of survival analysis in lymphoma cancer and the abilities of feature selection to solve the classification problems. Therefore, the aim of this paper is to propose PSO feature selection for the classification of survival analysis in lymphoma cancer. Experiment result from the proposed approach is then compared with the classification without feature selection. From the comparison results, classification of survival analysis with PSO feature selection outperformed classification of survival analysis without feature selection; with 85.4545% compared to 77.7726% each.