Visualized Implementation And Application Research Of PFCM Clustering Algorithm

Nowadays we have entered an era of information explosion. With thedevelopment of computer science and telecommunications, many traditionalindustries are now relied on digital and internet technology heavily. This brings out animportant topic: how can we discover valuable information from the huge amount ofdata, and reuse them in business. Data mining is the key to this problem and isbecoming a very popular research area of information technology.As one common topic in data mining, Data clustering is the prerequisite of thedata analysis. Data analysis is necessary when we want to get information orknowledge about the system that generates the data. However, in some cases theamount of data, or the system complexity, requires automatic methods to do this job.Besides, if the data is unlabeled, we need clustering methods to associate a label to asubset of data that are relatively close together. Data clustering is now extensivelyused among a wide range of industries and c-Means cluster algorithm is one of themost popurlar clustering methods.In this paper, we researches several c-Means cluster algorithms, includingHCM(Hard c-Means), FCM(Fuzzy c-Means) and PCM(Possibilistic c-Means), thenintroduces the PFCM (Possibilistic Fuzzy c-Means), of which the cluster distributionshave a better adaptation with the natural distribution of the data. The specificresearch contents includes: Introduction of the background and current status of theresearch, as well as the research purpose and meaningful value; Introduction forFuzzy Set theory as the background knowledge.This work is dedicated to the research of clustering algorithmns, and presents theimnprovemnents of the FCM and PCM algorithmn. Comnparative results that we gotwith different clustering algorithmns from data experiments are given by tables, followed by the discussion and a final conclusion. At last, c-Means clusteringmethods are applied in the research of the real estate market of35biggest cities inChina. Meaningful clustering results are got, which match the real word facts. Usefulconclusions are drawed regarding to the urban economics develpement status and andhousing price of these cities.