The Research Of Method Of Comprehensive Quality Assessment Of National Defense Students Based On Rough Set And SVM

With the increasing global competition, the importance of talent has been recognized by countries around the world. As an important basis for management's decisions in the department of human resources management, personnel assessment technology which can improve the efficiency of the management, is widely used in different companies and government agencies. However, how to improve the objectivity and accuracy of personnel assessment has been a difficult problem. The essence of the personnel assessment, which is a typical non-linear classification problem, is a comprehensive analysis to various indicators of data. Under the continuous development of computer technology, the knowledge and rule can be found from the data by the methods of machine learning. The methods of machine learning can guide the analysis of human judgments, thus been a lot of attention. The study of personnel evaluation by the methods of machine learning has become a hot research. Firstly, the comprehensive quality assessment indicators of national defense students have been built, and the quantify indicators and evaluation criteria have been determined. Secondly, analyses the basic principles of rough set theory, and introduce a simple way to reduce the quality attribute of national defense students. Thirdly, this paper describes the principle of support vector machine classification. After One-versus-Rest, One-versus-One, Binary tree SVMs and DAG-SVMs were analyzed, a Hybrid classification method which is based on the combination of binary tree SVMs and DAG-SVMs was proposed in this paper. Finally, rough set and support vector machines were applied to the comprehensive quality assessment of national defense students. And then, the rough set was used to simple the comprehensive quality attributes of the national defense students. The support vector machine was used to classification the comprehensive quality attributes of the national defense students. Making the national defense students' comprehensive quality indicators as input properties, the category of the national defense students were classified into excellent, good, qualified and problematic. In the classification experiment, the comprehensive quality assessment data of 170 national defense students is divided into two groups. The first group of data were collected in March of 2011, the distribution of training samples and test samples were divided by the ratio of 3 to 1. In the comprehensive quality assessment data of the second group, the data collected in March of 2011 was treated as training samples, and the data collected in May of 2011 was treated as testing samples.The experimental results show that the method of personnel assessment based on rough set and support vector machine can provide an objective basis for decision making by the manager. The use of the rough set can achieve the purpose of optimization to the training model. The classification accuracy of the method based on Rough sets and support vector machine is higher than the method based on SVM.