Effective criminal records of hate crimes in Canada using clustering - A comparison of EM and simpleKmeans

• 발행기관: 한국IT마케팅학회
• 저자명: 김소희(So Hee Kim) 정용규(Yong Gyu Jung)
• 간행물 정보: 『한국IT마케팅학회 논문집』2015 IT마케팅학회 논문집 Vol.1 No.1, 140~143쪽, 전체 4쪽
• 주제분류: 경제경영 > 경영학
• 파일형태: PDF
• 발행일자: 2015.11.20

영문 초록

Not only South Korea, there is no one place to be ill the head for crime in all countries of the world, including Canada. In this paper, With an emphasis on criminal record, we are focused on how to effectively its record. Furthermore, This data is that it is how predictable the crime based on, I think I'll try to study as to how to be used in real life. To try analyzed using the EM algorithm and SimpleKMeans way its research methods. Thus, in order to solve the problem of crime, it would use data mining techniques. Accordingly, to try to evaluate the navigation algorithm that greatly affects the performance experimentally. As a result, a number of algorithms, such as the EM algorithm and SimpleKMeans method is expanded, it was proposed. And for certain situations, it has become to require the procedure for assessing the suitability of algorithms for each situation that It was based on the verification of the performance experiment and results to selectively use the best algorithm. In this paper, using the same data for the EM algorithm and SimpleKmeans algorithm, based on a result of cross-validation to evaluate the performance of both algorithms, Especially in the navigation algorithm that greatly affect the performance of a cluster analysis. Thereby was analyzed on the basis of the application result changes with experimental data in performance due to enlarged navigation algorithms. And by applying the resample filters were described what kind of differences before.

목차

Abstract

I. Introduction

II. Research Literature

III. Experiment

IV. Results and Evaluation

References

키워드

EM simpleKmeans A hate crime