Font Size: a A A

Violent Crime Characteristics Of Clustering Analysis And Application Based On Improved FCM Algorithm

Posted on:2014-01-14Degree:MasterType:Thesis
Country:ChinaCandidate:X L DingFull Text:PDF
GTID:2248330398952531Subject:Computer Science and Technology
Abstract/Summary:PDF Full Text Request
In years of work of our country’s Public security bureau, the officials promote Informatization continually on one hand,and It has a considerable scale data accumulation of public security work special data and social informatization. Data mining has become a important and meaningful topic for analyzing all kinds of criminal motive.Because of most of problems has happened is a result of the interactions among different subjects, so it has become a difficult problem for tacking with data of interdiscipline. As a kind of data mining, Cluster Analysis could do a lot of analysis for mass data and get valuable information from them, and applied in data mining, artificial intelligence. Pattern Recognition and fault diagnosis successfully. Cluster Analysis also applies to tackle with data of interdisciplinary well. The purpose of this paper is analyzing criminal data by using Cluster Analysis method, obtain valuable information from the criminal background information, psychological information and genetic information comprehensive mining, in order to found the factors that affect and cause crime. Cluster Analysis could help public security personnel to targeted education of criminals and trace its origin influence factors for preventing c rime.Finally, Cluster Analysis convert the advantage of National public security organs informatization to the effective Policing abilities. Improve National public security’s ability of preventing or fighting crimes.This paper based on the analysis of relevant research, researching and analyzing traditional hard clustering algorithm and fuzzy clustering method, and according to the characteristics of the violent crime data, raised the fuzzy clustering method based on continuous attributes discretization. This paper tries to put the information entropy discretization data preprocessing algorithm applying data pre-processing, In order to achieve better clustering effect.At last public security organ could use the algorithm of this paper raised applying to the violent crime in factor analysis, tries to analyze the deep reasons of the violent crime and predicts violent crime, the experimental results are analyzed in details.
Keywords/Search Tags:Fuzzy clustering, The discretization of continuous attributes, Information Entropy, Analysis of the factors of violent crime
PDF Full Text Request
Related items