Control Systems and Computers, N6, 2016, Article 8
DOI: https://doi.org/10.15407/usim.2016.06.067
Upr. sist. maš., 2016, Issue 6 (266), pp. 67-72.
UDC 343.346.8:004.056.53
E.V. Bodjanskiy, Doctor of Sci. (Eng.), Kharkiv National University of Radio Electronics, Ukraine, Nauky Ave. 14, Kharkiv, E-mail: bodyanskiy@gmail.com,
V.M.Strukov, Ph.D. (Eng.), Kharkiv National University of Internal Affairs, Ukraine, E-mail: struk_vm@ukr.net,
D.J. Uzlov, Ph.D. (Eng.), Office of Information Support of the Main Directorate of the National Police in the Kharkov region, head of the department, Ukraine, E-mail: poputchik@і.net
The Task of the Proximity Estimation of Multidimensional Objects of the Data Analysis
Introduction. The task of the proximity estimation of multidimensional objects is well investigated [1,2]. But there are some tasks, which have features that enable to apply the classic algorithms and methods. Such practically significant task is processing of multidimensional objects with different measurement scales properties that are stored at data base of information departments of law enforcement of Ukraine.
Purpose. Investigation purpose is to develop the way foregoing objects proximity estimation to enable applying classic methods of clustering, classification and association.
Methods. The approach to formalization of data array features is proposed and expressions for numerical, rank and complex metrics in multidimensional objects space with the properties in the different measurement scales are developed.
Results. The developed approach enables applying classic methods of clustering, classification and association for data base of information departments of law enforcement of Ukraine processing, in particular, for solving the significant problem of the detection of the implicit and hidden relations between criminal accounting objects in data bases of information systems of law-enforcement agencies.
Download full text! (In Russian)
Keywords: Data Mining, multidimensional objects, clustering, classification, measurement scales, numerical metric, categorial metric, rank metric.
1. Han L., Kamber M. Data Mining: Concepts and Techniques. – Amsterdam: Morgan Kaufman Publ., 2006. 754 p.
2. Aggarwal C.C. Data Mining. Cham: Springer Ltd. Publ. Switzerland, 2015. 734 p.
3. Hathaway R.J., Bezdek J.C. Fuzzy c-means clustering of incomplete data. IEEE Trans. On Systems, Man and Cybernetics. 2001. 31 (5). P. 735–744.
4. Brouwer R.K. Fuzzy set covering of a set of ordinal attributes without parameter sharing. Fuzzy Sets and Systems. 2006. 157 (13). P. 1775–1786.
5. Pedricz W., Chen Sh.-M. Information Granularity, Big Data and Computational Intelligence. Cham: Springer, 2015. 444 p.
6.Instruktsiya pro yedynyy oblik zlochyniv.http://zakon4.
rada gov.ua/laws/show/v0020900-02/page.
7. Metodychni rekomendatsiyi shchodo alhorytmu diy korystuvachiv z orhanizatsiyi formuvannya Intehrovanoyi informatsiyno-poshukovoyi systemy orhaniv vnutrishnikh sprav Ukrayiny. vid 16.01.2014 № 727/Z. – K.: MVS Ukrayiny, 2014. 35 p.
8. Westphal C. Data Mining for Intelligence, Fraud and Criminal Detection. Advanced Analytic & Information.
Received 23.11.2016
