Staff Turnover Warning Information System Based On Decision Tree

Employees as the creator of the inheritance and the value of the enterprise achievements, to enterprise's survival and development plays an important role, is the enterprise of precious wealth. But "the old and new change frequently " are common phenomenon in the enterprise, and the loss of employees will inevitably bring irreparable loss, as a result, many companies increasingly attaches great importance to the staff turnover problem. This article in view of the company S staff turnover problems existing in human resource management, on the basis of the research on related literature, analyzes how to prevent losing staff in advance, and put forward the concept of staff turnover warning information system, how to build the classification problem can be converted to early warning model and early warning information system analysis and design.First of all, through the contrast research of mainstream classification algorithm, it is concluded that this study used the decision tree classification model. Secondly, staff turnover of possible factors influencing the binomial Logistic regression analysis, and erosion is a significant correlation factors, reference attribute as building early warning model. Finally, analysis how to build the early warning model, and carries on the forecast accuracy of the model, the resulting warning rules, provide theory support for the system to achieve early warning function.Demands for staff turnover warning information system through detailed analysis, have been clear about the early warning information system function, performance, and business requirements, technical feasibility and operation feasibility of system are discussed. And through the integration of DMX, decision tree model was applied to information system, realize the early warning information system based on BIDS. In this paper, on the basis of demand analysis, staff turnover warning information system for the detailed summary design, detailed design and database design, and tries to design a reasonable and efficient information system. Profile design will demand analysis of system use case diagram convert data structure, software structure and flow chart of early warning system architecture diagram and the loss is obtained. The detailed design to solve each module of the control process, the internal data structures and algorithms. Then the database concept structure and logical structure design was carried out. Finally, made a summary of this topic research, and points out the shortage and the late research content. In this paper, a total of five parts, each part of the content, respectively, as follows:Chapter one is the introduction part introduces the background of the selected topic and the significance, the domestic and foreign scholars about the staff turnover and the application of decision tree related literature were reviewed, at the same time elaborated the research content and research framework.The second chapter of this study involves the related theory. First the distinction is made for staff turnover and flow, analyzes the possible factors influencing the loss of; Second, the classification algorithm has carried on the contrast analysis, determine the study of the classification methods, decision tree classification; Finally expounded how to classification algorithm was applied to staff turnover warning- loss of early warning information system.Chapter three on the early warning information system has carried on the detailed demand and feasibility analysis, clear the function of the system, the performance and business requirements, discussed the technical feasibility and operating system. Building was determined by the loss factor regression analysis, classification model reference attribute, and expounds the logic of erosion classification model.Chapter iv as the focus of this study, starting with the theoretical mechanism of early warning model, expounds on the early warning model building process, the resulting warning rules, and according to the results analyzed the impact of the current company S staff turnover of the key factors, put forward the corresponding Suggestions. Finally analyzes how the early warning model is applied in information system, to early warning of staff turnover, namely the DMX integration design method library and based on the rules of early warning system.The fifth chapter mainly solves the problems of early warning system for how to do. According to the demand analysis of early warning information system has carried on the outline design, detailed design and database design, tries to design a reasonable and efficient system.The characteristics of this paper is to: 1) the decision tree classification in solving staff turnover in the field of application has carried on the theory discussion, put forward the concept of loss of early warning information system; 2) by integrating DMX combines the decision tree model and information system, realize the staff turnover warning.Enterprise by means of the effective early warning information system of recruitment and existing employees classified forecast, early warning of staff turnover, improve the stability of the team, for the enterprise provides a reference for solving the problem of staff turnover.