Home // SMART 2019, The Eighth International Conference on Smart Cities, Systems, Devices and Technologies // View article

Predicting Incidents of Crime Through LSTM Neural Networks in Smart City Domain

Authors:
Ulises M. Ramirez-Alcocer
Edgar Tello-Leal
Jonathan A. Mata-Torres

Keywords: LSTM; Prediction; Neural Network; Data Classification; Smart City;

Abstract:
Crimes are common social problems that affect the quality of life, economic growth and reputation of a country. In smart cities, the aim is to reduce crime rates using Information and Communication Technologies (ICT), specifically with the use of Internet of Things (IoT) technology in combination with legacy information systems, in order to obtain data automatically. In this paper, we propose an approach based on deep learning for the classification of incidents of a crime of public safety through predictive analysis. The predictive model is based on a neural network Long Short-Term Memory (LSTM), trained with a small group of attributes, enabling the prediction of the class label in the validation stage, with a high percentage of prediction accuracy. The proposed approach is evaluated through a big data set (real data) of type open data, which contains historical information about the crimes of a smart city.

Pages: 32 to 37

Copyright: Copyright (c) IARIA, 2019

Publication date: July 28, 2019

Published in: conference

ISSN: 2308-3727

ISBN: 978-1-61208-730-6

Location: Nice, France

Dates: from July 28, 2019 to August 2, 2019