Advertisement

STL-FNN: An Intelligent Prediction Model of Daily Theft Level

  • Shaochong Shi
  • Peng ChenEmail author
  • Zhaolong Zeng
  • Xiaocheng Hu
Conference paper
  • 5 Downloads
Part of the Advances in Intelligent Systems and Computing book series (AISC, volume 1143)

Abstract

Theft is a long-standing crime against property, which exists alongside the development of private ownership. If the time pattern of thefts could be found out well, police officers can take preventive measures in advance to control the occurrence of theft crimes. However, thieves are smart enough to pick targets randomly. This makes it difficult for police to predict accurately when and where the offenders would commit crimes. In this paper, STL-FNN is proposed based on a seasonal trend decomposition procedure based on loess (STL) and full-connected neural network (FNN),which is designed to predict daily theft level in order to find out when criminals are most likely to commit crimes. The empirical case of prediction of daily theft level in City B shows that the STL-FNN model is better than the other five traditional models for prediction of the long-term sequence (365 days). This model is expected to have high potential application value in the schedule of anti-theft activities planned by polices.

Keywords

Crime prediction STL-FNN Daily theft level 

Notes

Acknowledgements

This work was supported by Natural Science Foundation project (71704183) and Beijing Natural Science Foundation (9192022). Also, it is grateful to the sponsorship from foundation of National Engineering Laboratory for Public Safety Risk Perception and Control by Big Data and National Engineering Laboratory for Public Safety Risk Perception and Control by Big Data.

References

  1. 1.
    Song, L.G., Zhang, L.F.: a new interpretation of the crime of theft stealing mode. Hebei Law Sci. 33(8), 94–100 (2015)Google Scholar
  2. 2.
    Huang, C., Li, J.H.: Method of crime prediction. J. JiangSu Police Off. Coll. 26(1), 107–110 (2011)MathSciNetGoogle Scholar
  3. 3.
    Grzegorz, B., Zbigniew, M.W., Paweł, C.: Time series analysis for crime forecasting. In: 26th International Conference on Systems Engineering, pp. 1–10. IEEE, Sydney, Australia (2018)Google Scholar
  4. 4.
    Chandra, B., Gupta, M., Gupta, M.P.: A multivariate time series clustering approach for crime trends prediction. In: 2008 IEEE International Conference on Systems, Man and Cybernetics, pp. 892–896. IEEE, Singapore (2008)Google Scholar
  5. 5.
    Yadav, R., Kumari Sheoran, S.: Crime prediction using auto regression techniques for time series data. In: 3rd International Conference and Workshops on Recent Advances and Innovations in Engineering, pp. 1–5. IEEE, Jaipur, India (2018)Google Scholar
  6. 6.
    Xie, J.H.: Time series prediction based on recurrent LS-SVM with mixed kernel. In: 2009 Asia-Pacific Conference on Information Processing, pp, 113–116. IEEE, Shenzhen, China (2009)Google Scholar
  7. 7.
    Chen, P., Hu, S.Y., Luo, W.J.: Study about crime reporting forecasting based on Grey-Markov Model. China Public Security. Acad. Ed. 2, 32–35 (2014)Google Scholar
  8. 8.
    Tu, X.M., Chen, G.Q.: A hybrid ARIMA-LSSVM model for crime time series forecast. Comput. Technol. Appl. 41(2), 160–163 (2015)Google Scholar
  9. 9.
    Liu, M.L., Gao, J., Huang, H.Z.: A model of crime intelligence prediction based on a hybrid model of spatio-temporal sequence. J. Intell. 37(9), 27–31 (2018)Google Scholar
  10. 10.
    Cleveland, R.B., Cleveland, W.S., McRae, J.E.: STL: a seasonal_trend decomposition procedure base on loess. J. Off. Stat. 6(1), 3–73 (1990)Google Scholar
  11. 11.
    Brett, L.: Machine learning with R. PACKT Publishing Ltd, Birmingham, UK (2016)Google Scholar

Copyright information

© Springer Nature Singapore Pte Ltd. 2021

Authors and Affiliations

  1. 1.Department of Information Technology and Cyber SecurityPeople’s Public Security, University of ChinaBeijingChina
  2. 2.Key Laboratory of Security Technology & Risk AssessmentMinistry of Public SecurityBeijingChina
  3. 3.National Engineering Laboratory for Public Safety Risk Perception and Control by Big DataBeijingChina
  4. 4.China Academy of Electronics and Information TechnologyBeijingChina

Personalised recommendations