Please use this identifier to cite or link to this item: http://hdl.handle.net/20.500.12188/22846
Title: Development of regression models for national crime forecasting
Authors: Petroski, Nikola
Dedinec, Aleksandar
Filiposka, Sonja 
Mishev, Anastas 
Issue Date: Oct-2021
Conference: 6th Working Group Meeting DIGFORASP Cost Action, Budapest, Hungary
Abstract: Reducing national crime rate is an extremely important, but also difficult problem. The goal of this research is to look for and discover patterns in the crime occurences and the spatial and temporal parameters connected to them, therefore making it possible to predict future crime rates based on those parameters. We use data from the year 2011 until now through a system which reads daily publications from the official website of the Macedonian Ministry of Interior, written in natural language, from which information about the date, location, type of crime and description are extracted. Since the accuracy of this database is crucial for the precision of the crime analysis and forecasting, an additional data verification and cleaning process was conducted, certain inconsistencies were corrected and additional detailed information about the municipality and settlement of the crimes was added. We set up the crime analysis and forecasting as a regression problem, predicting the number of crimes that would happen on a given date in a given location, knowing the statistics about the days before that. From there, we compared the results with another, classification based crime forecasting study made on the same region to find out in which ways the parameters influence the end result, and how we can improve that.
URI: http://hdl.handle.net/20.500.12188/22846
Appears in Collections:Faculty of Computer Science and Engineering: Conference papers

Files in This Item:
File Description SizeFormat 
presentation (3).pdf179.65 kBAdobe PDFView/Open
Show full item record

Page view(s)

16
checked on Jan 28, 2023

Download(s)

4
checked on Jan 28, 2023

Google ScholarTM

Check


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.