Method Study Of Human Resource Selection Based On Support Vector Machine

Human resource selection is a basic step in human resource management and directly affects training, performance, payment and other activities. There are two important processes in human resource selection. One is using measurement tools to gather information from applicants; the other is evaluating this information to make selection decision.But current selection methods are far from perfection, especially decision-making methods. Traditional judgmental methods are influenced by subjectivity and statistical methods aren't fit for small samples. So these methods cannot effectively guide our practice.According to such situation, this thesis introduced a new progress in data mining—support vector machine (SVM). SVM is based on statistical learning theory, which is designed for small samples. In addition, we combined SVM with rough set theory, and got rough support vector machine algorithm (RSVM). RSVM can automatically find key attributes which affect performance and build performance prediction model by analyzing data. And thus effectively avoid subjectivity. At last a new personnel selection model based on RSVM was proposed, and the experiments showed that the new model achieves a better accuracy than the existing ones.