Prediction of crime from time series data-driven model

Grzegorz Borowik , Zbigniew M. Wawrzyniak , Paweł Cichosz , Radosław Pytlak , Eliza Szczechla , Paweł Michalak , Dobiesław Ircha , Wojciech Olszewski

Abstract

The paper presents a short-term prognosis of the crime rate for subsequent two years in a selected region of Poland. The estimate of the future crime rate has been developed as data-driven model on the basis of past crime time series covering the period of three years. In our research we have focused on certain types of crime. The vast majority of crime categories experienced a downward trend in recent years, therefore the majority of forecasts assume further decline. The methods used for prognoses combine time series regressive methods with generalized additive models.
Author Grzegorz Borowik
Grzegorz Borowik,,
-
, Zbigniew M. Wawrzyniak (FEIT / PE)
Zbigniew M. Wawrzyniak,,
- The Institute of Electronic Systems
, Paweł Cichosz (FEIT / IN)
Paweł Cichosz,,
- The Institute of Computer Science
, Radosław Pytlak (FMIS / DAICM)
Radosław Pytlak,,
- Department of Artificial Intelligence and Computational Methods
, Eliza Szczechla
Eliza Szczechla,,
-
, Paweł Michalak
Paweł Michalak,,
-
, Dobiesław Ircha
Dobiesław Ircha,,
-
, Wojciech Olszewski
Wojciech Olszewski,,
-
Pages1554-1565
Publication size in sheets0.3
Book Valenzuela Olga, Rojas Fernando, Pomares Héctor, Rojas Ignacio (eds.): ITISE 2018 International Conference on Time Series and Forecasting, Proceedings of papers, vol. 1, 2018, Godle Impresiones Digitales S.L., ISBN 978-84-17293-57-4, 570 p.
Keywords in Englishcrime prediction, time series, generalized additive model
Languageen angielski
File
borowik_prediction.pdf 1.7 MB
Score (nominal)0
Citation count*
Cite
Share Share

Get link to the record


* presented citation count is obtained through Internet information analysis and it is close to the number calculated by the Publish or Perish system.
Back