Density Algorithm And Its Application In Hrm Research

From the view point of economics, to obtain the postgraduate degree belongs to a With the rapid development of modern enterprise,data generated from different information systems become more and more.It is really not easy to extract useful information from such amount of sources.How to utilize the huge original data to analyse curren situation and predict future of quantities effectively,has already become a great challenge that the human beings have faced.Data Mining is devised to solve the problem.Cluster analysis is an important research area in datamining.Nowadays,With the technology development such as satellite remote sensing, sensor networks, high-energy physics research, Large amount of data is stored in the database,which have the characteristics of High dimension, sparse distribution and many noise data.In many applications,these datas distribute on different nodes.If you want to get information from these distribute datas using the method of traditional algorithm,you must merger these datas to the same center.Because of the constraints of transmission speed and security,It is difficult to concentrate the data of each site to the center site.In some areas,it is almost impossible to concentrate the data set to one site,the additional cost is large.K-Dmeans algorithm is based on the Distributed Clustering Algorithm of K-Means,this paper directting the lacks imoproved a lot, such as K-Dmeans algorithm need to send many data objects in every iterative process, because of bandwidth limitations, network latency and other issues cause a large communication cost,especially in dealing with large data sets,the communication cost is far greater than the cost of computing,algorithm is very low ets..The improved algorithm can effectively deal with transmission of a small amount of clustering information,the efficient of implementation is very high.It also can sovle the problems of the distributed density algrithom,such as weak ability in the noises and abnormal datas,not suiting to high dimensional data and the lacks of local node'clustering result is large etc.In the face of applying theory technology to practice,this paper analysis chinese enterprise'performance evaluation system of the existing status as well as methods of performance evaluation method,With the experiences that the author's developing the system of HRM,a new performance evaluation method is put forward in this paper.This method is an employee performance evaluation clustering model,which based on improved density clustering algorithm. The detail method is that: At first,according to every kind of employee's performance evaluation index,the employee performance evaluation clustering model is established.Then the employees are clustered and classified by using improved density clustering algorithm.And the classification result can offer decision support for personnel program and adjusting.In this article,the author first introduces distributed clustering algorithm research, Performance Management Development,and cluster analysis,development,and cluster analysis methods and the application of cluster analysis,then it introduces distributed clustering algorithm based on density,and principle of distributed clustering algorithm, the strengths and weaknesses of distributed clustering algorithm analysis, and set out several improvements in existing methods.Based on above theories,the new clustering analysis method,which based on Improved density algorithm,is put forward in this paper.Then the author realizes the Improved Density Algorithm and design an emulator.With the emulator,the algorithm is compared with other clustering analysis algotithms.Then, In chapter 3 the author introduces the human resource management theries and expatiates on the fixed position of the performance evaluation in human resource manage firstly,and then analyses the advantages and disadvantages of the curren performance evaluation methods.After that,the integrant knowledge of the employee performance evaluation clustering model is expatiated.In chapter 4,the author realizes a complete employee performance evaluation system.According large numbers of data test and result analysis,it is aproved that the Improved Density Algorithm which is based on local density k -PCLDHD to get clustering and noise points,then use K-Means to generate all the special centers,and past these special centers and noise points to all the slave sites,can solve high-dimensional data and the questions of transmission large number points and so on. Finally,this paper summaries the whole work and puts forward the further work and expectation.