| The population issue has been the greatest concern of the society as a whole, thepopulation migration flows and their impact factors and management practices is themost intractable problems of the relevant departments of the management workers. Inthe study of population issues in the past, there have been a lot of data, but these datain population management and more managers directly used to query or statistics,implied by these data we need mostor the most useful information for us but we havenot been found. Data mining technology just to solve the above problem, because datamining is massive and there is no direct law, the main purpose of this process is a lotof data mining of our most valuable information, and data mining clustering andassociation rule mining and its application is very important and valuable part of thedata mining area. This article first introduce in the relevant sections of the basicconcept of data mining, and focused on the clustering methods in data mining andassociation rule mining and mining algorithms and realization.In the first part, focuses on the research background and current situation of thearticle, summarize by discussing the research on population changes on the floatingpopulation or youth, the elderly and other special age of the population, given themore practical needs. selected regional changes in population, but the whereabouts ofthe floating population survey does not reflect the ’region’ we are asking for this object,so we based on the same point to the region this in order to migrate the differencebetween the rate of population and net migration the data itself as the goal to conducta study to determine a specific object.In the second part, the authors first GIS technology, to speak to the data itself, thedifference between the rate of net migration of various cities and economic factorsreflected visually on a map, and strive to be simple and clear, while for the next Thebasis for the study.In the third section, following the second part, with the intuitive realization of theobject of study data and the problems raised, the following discussion is to cluster the data of21cities object in Sichuan Province, and select center distance clusteringalgorithm, the tool of choice in the algorithm, taking into account in this part of the21objects is not too much data, if the complex algorithms or use large tools to obtain theresults but will probably not ideal, the authors boldly chose a convenient SPSS tools.In the fourth section, the difference between the rate of net migration of the firstpart of the various cities in Sichuan Province, migration of population distributioncharacteristics and conclusions as the basis for new elements to the second part of thecluster in the part, data of various cities in Sichuan Province, the economic andnon-economic factors the introduction of select classic AP iterative algorithm andoracle to explore the21to Fen are different types of associated factors. And make abrief analysis of the results, then proposed population management policy. |
