Human Resource Selection Based on Performance ClassificationUsingWeighted Support Vector Machine
Qiangwei Wang, Boyang Li, Jinglu Hu · Journal of Advanced Computational Intelligence and Intelligent Informatics · 2009
Recruitment and selection have the first priority in human resource management. Poor selection decisions can be enormously costly to organization. So a valid decision method is needed. Traditional selection method is based on linear model. However, it is not proper for this intricate nonlinear relationship between applicants and job performance. In this paper, we introduce a new human resource selection system using weighted support vector machine (WSVM) which is fit for the nonlinear problem. It gives the selection process a feedback and improves classification results. Besides, we also proposed a new weight generating method to keep the characteristic of human resource. It reduces the effect of outliers and noise for classification, and distinguishes different importance of selection criterions. Furthermore, different weights are compared for WSVM. Questionnaire survey was issued to acquire dataset. Simulation results show that our proposed selection system is valid for human resource selection; it performs higher classification accuracy than traditional linear method. It can be used to support the decision making in human resource selection. Besides, different weight generating methods are compared for WSVM, our proposed weight generating method obtains better efficiency than standard SVM and class center based weight generating method.