Application Research On Data Mining Technology In The University Human Resources Scale

The21st century, with the scale of college continues to expand, then bring the teacher resources tension, explores the resource utilization rate is low, the college education funding problem and so on related problems. Along with the development of information technology, college has accumulated a large number of relevant university data, through the use of efficient data mining technology to a large number of university human resources related to the historical data analysis, there are correlation value relations and laws.First of all, the data mining related concepts are narrated. Briefly we introduces the data mining technology is commonly used artificial neural network, genetic algorithm, the decision tree algorithm and so on. It also introduces the related concepts of data mining system. This paper mainly introduces the structure of BP network, learning algorithm and the algorithm process, the limitation and improvement measures, etc. The BP neural network is applied to college student/teacher ratio model validation and forecast. The paper focus on the optimization algorithm of genetic algorithm and particle swarm optimization algorithm learning process and structure design. In order to optimize the BP neural network laid the foundation.Secondly, this paper research university human resources scale as the starting point, this paper introduces the university human resources student/teacher ratio in the related concept, collect the Liaoning province nearly more than ten years of university human resource related data, using the software MATLAB7.0to the university student/teacher ratio model building, in the modeling of data pretreatment prior to, including data of the dimensional change, normalized processing, principal component analysis and other related work. Through a lot of related experimental verification, the BP network model to determine the number of neurons in hidden layer, and finally determine the rational structure of the training of network. This paper studies the comprehensive, normal, national colleges and universities; Engineering, agriculture, forestry colleges and universities; Medical colleges and universities three categories of college student/teacher ratio. And the three types of universities student/teacher ratio index verification and analysis. And the genetic algorithm to optimize the BP neural network and particle swarm optimization model based on the BP neural network used in three universities in the student/teacher ratio. Through,comparative analysis of the model prediction accuracy and relative error.Finally, The paper introduced the related prediction model method, the university student/teacher ratio using the rolling prediction method to predict the next few years college student/teacher ratio variation trend. Finally, According to the prediction of university student/teacher ratio to the analysis of the next few years the size of the university human resources and the utilization ratio of the situation.