摘要
公安部门大量案件物证信息堆积,相关部门只能进行基础性查询,案件物证信息使用职能单一。基于K-means算法案件信息分类、预测,可以为相关部门提供额外的信息汇总,有效打击犯罪。通过案件信息中案件类型、案件发生地域等特点,建立相应的模型。利用K-means算法简洁、高效、容易实施的特点,预测不同地域不同案件发生的概率。论文在传统的K-means算法缺陷及其论证上,借鉴其他改进算法,通过实验证明,该改进算法可以更为准确地为案件分类,进而反映案件真实的发生状况。
Public security departments have a large number of cases material information,relevant departments can only conduct basic inquiries,evidence information of cases using a single function.Classification,forecasting of evidence information based on the K-means,which can provide additional information of summary for relevant departments,effectively combat crime.Through case information characteristics like case type,the occurrence of cases and many more,establish the appropriate model.Using the features of K-means algorithm like simple,efficient,implement easily,probability of occurrence of different cases in different regions can be predicted.In this paper,in view of the defects of classic K-means algorithm and some related arguments and then learn from other improved algorithms,the experiment shows that the improved algorithm can be more accurate for the classification of cases,and then response to the real situation of the case,proved by experiment.
作者
王健豪
苏勇
WANG Jianhao;SU Yong(School of Computer,Jiangsu University of Science and Technology,Zhenjiang 212003)
出处
《计算机与数字工程》
2019年第8期1999-2001,2032,共4页
Computer & Digital Engineering
