The Algorithm Research Of The Target Group Distribution Characteristics

The target group distribution characteristics of the analysis is widely used, have playedan important role in many areas. In the target group distribution characteristics of the analysismethod, cluster analysis is one of the most commonly used algorithms. Traditional clusteranalysis is generally based on the prototype of cluster centers, which can not identify clustersof arbitrary shape, and most of the clustering algorithm during clustering are required to givesome a priori parameters, such as the FCM algorithm predetermined poly class and thenumber of categories C, but due to the distribution range of the target group showed a randomdistribution, usually there is no fixed distribution shape, in order to obtain such priorinformation is very difficult. Therefore, the traditional cluster analysis algorithm is not anaccurate distribution characteristic of the target group for analysis.In this paper, the research focuses on the distribution characteristics of the target group.Based on the relationship between graph theory analysis methods from point to point andpoint to point, you can simplify the many complex issues, and can be a good distinction toidentify the various shapes of graphics. This article uses cluster analysis algorithm based ongraph theory to analyze the distribution characteristics of the target group. Simulation resultsshow that the algorithm can described in the extraction of the distribution parameters of thetarget group well on the basis of the distribution characteristics of the target group.This article also uses a description method based on Fourier descriptors of shape, theregional boundaries of the target group use Fourier descriptors to describe the distribution ofthe target group to get the shape information. The description method has the characteristicsof displacement, rotation, size, starting the same nature and amount of data is small, has goodapplicability to describe the shape of the target group characteristics. The simulation resultsshow that the algorithm has a strong shape recognition capability, enable to describe the shapecharacteristics of the distribution of the target group.