Open AccessMethods of data to determine the class of the dependent variable in the analysis of murders in Kiev with sharp objects
Author(s) -
Volodymyr Zosimenko,
Hanna Zolnikova
Publication year - 2016
Publication title -
sudovo-medična ekspertiza
Language(s) - English
Resource type - Journals
eISSN - 2707-8728
pISSN - 2218-838X
DOI - 10.24061/2707-8728.1.2016.4
Subject(s) - object (grammar) , statistical analysis , class (philosophy) , computer science , variable (mathematics) , research object , data science , statistics , data mining , mathematics , artificial intelligence , geography , regional science , mathematical analysis
During 2011-2014 years we conducted a retrospective study of 849 non-selective deaths in Kiev from wounds, that were caused by a sharp object. The aim of this study was to determine the optimal algorithms of the statistical processing of a data, as a tool to justify the results of the analysis of cases of murders using sharp objects in Kiev.
Materials and methods. Generally, the object of the study were 1068 “Acts of the forensic investigation of corpses” and “expert opinion” about the killings with using a sharp object, obtained from the archives of the Kiev City Bureau of forensics. Depth study was 15 years (1997-2011 years). Among them - 930 cases of the availability of data on the offender (age, gender), of which we selected 849 cases, in which one offender met one victim (corpse).
Results. The article presents selected our approaches to the definition of statistical data processing algorithms as a tool to support the analysis of murders with sharp objects.
Conclusion. We have determined the optimal algorithm such statistical data - the use of technology Data Mining, which will continue to be applied for the construction of the system, which is known for filling elements of criminalistics characteristics of crime, the characteristics of the offender will get a certain statistical confidence.