An Early Warning Method Of Brain Drain In High-tech Enterprise Based On SVM

With the continuous development of modern society, the competitions among enterprises are becoming more and more fierce. Loyalty of the staffs is definitely a great challenge of the modern companies which are mainly high-tech and specialists oriented. Therefore, how to keep the staffs loyal by using the "intellectual alarm system" with the minimum cost is obviously has significant meaning.This article based on the daily performance record of the high-tech enterprise staffs in HR department. The survey questionnaires were designed for the staffs that work shortly in such companies with relatively young age and higher education degree. All the date I received have been analyzed and categorized by using the support vector machine. Meanwhile, setting up a model that has stable characteristics is necessary to forecast the potential data and category by analyzing the previous data.Support vector machine applies well on an early warming method of brain drain based on SVM classification. We can see from the experiment result that the accuracy of choosing the Gaussian Function is good. In addition, support vector machine works practically for the small sample of Chinese high-tech enterprises. This model can alarm the brain drain with high accuracy. Moreover, it offers technical support to the manager for judging the staffs and making clever decision.In the last chapter of this article, some suggestions have been given to prevent the brain drain pheromone. Through external factors and internal personnel situation analyze the results of the brain drain, make an effective plans for C&B management, establish a performance evaluation management system based on incentive person, establish a comprehensive staff development system, establish a job rotation system, establish a work-sharing mechanisms and work-leaving communication. Hopefully it helps to deal with the current brain drain problem in high-tech enterprises.