Font Size:
A Spatio-Temporal Mixture Model for Urban Crimes
Last modified: 2018-05-28
Abstract
This paper considers the determinants of severe crimes at the censustract
level in Pittsburgh, Pennsylvania. We develop a mixture panel data model to
describe the number of severe crimes that allows for temporal as well as spatial
correlation, together with significant heterogeneity across census tracts. We use traditional Bayesian mixtures admitting uncertainty about the number of groups. We
focus on pooling regression coefficients across clusters, implying that census-tracts
belonging to the same cluster are similar. The clustering is done in a data-based
fashion.
level in Pittsburgh, Pennsylvania. We develop a mixture panel data model to
describe the number of severe crimes that allows for temporal as well as spatial
correlation, together with significant heterogeneity across census tracts. We use traditional Bayesian mixtures admitting uncertainty about the number of groups. We
focus on pooling regression coefficients across clusters, implying that census-tracts
belonging to the same cluster are similar. The clustering is done in a data-based
fashion.
References
1. Felson, M., Clarke, R.V.: Opportunity Makes the Thief: Practical Theory for Crime Prevention. In: Police Research Studies. London (1998)
2. Gelman, A.: Inference and Monitoring Convergence. In Gilks, W.R and Richardson, S. and Speigelhalter, D. (eds). Chapman & Hall:Boca Raton; 131-143 (1996)
3. Gilks, W.R. and Wild, P. : Adaptive Rejection sampling for Gibbs Sampling. Journal of the Royal Statistical Society, Series C, 41, 337–348 (1992)
4. Ishwaran, H., James, L.: Approximate Dirichlet process computing in finite normal mixture: Smoothing and prior information. J. Comput. Graph. Stat. 11, 508–532 (2002)
5. Kikuchi, G.: Neighborhood Structures and Crime: A Spatial Analysis. El Paso, Texas (2010)
6. Korobilis, D., Gilmartin, M.: On regional unemployment: An empirical examination of the determinants of geographical differentials in the UK. Scott. J. Polit. Econ. (2012) doi: 10.1111/j.1467-9485.2011.00575.x
7. Liesenfeld, R., Richard, J.F., Vogler, J.: Likelihood-based inference and prediction in spatiotemporal panel count models for urban crime. J. Appl. Econ. (2016) doi: 10.1002/jae.2534
8. Manolopoulou, I., Chan, C., West, M.: Selection sampling from large data ses for targeted inference in mixture modeling. Bayesian Anal. 3, 429–450 (2010)
9. Paap, R., Franses, P.H., van Dijk, D.: Does Africa grow slower than Asia, Latin America and the Middle East? Evidence from a new data-based classification method. J. Dev. Econ. 77, 553–570 (2005)
10. Ratcliffe, J.H.: Crime mapping: spatial and temporal challenges. Handbook of Quantitative Criminology, Piquero, A.R., Weisburd, D. (eds). Springer: New York, 524 (2013)
11. Roth, R.E., Ross, K.S., Finch, B.G., Luo,W., MacEachren, A.M.: Spatiotemporal crime analysis in U.S. law enforcement agencies: current practices and unmet needs. J. Gov. Inf. 30, 226–240 (2013)
2. Gelman, A.: Inference and Monitoring Convergence. In Gilks, W.R and Richardson, S. and Speigelhalter, D. (eds). Chapman & Hall:Boca Raton; 131-143 (1996)
3. Gilks, W.R. and Wild, P. : Adaptive Rejection sampling for Gibbs Sampling. Journal of the Royal Statistical Society, Series C, 41, 337–348 (1992)
4. Ishwaran, H., James, L.: Approximate Dirichlet process computing in finite normal mixture: Smoothing and prior information. J. Comput. Graph. Stat. 11, 508–532 (2002)
5. Kikuchi, G.: Neighborhood Structures and Crime: A Spatial Analysis. El Paso, Texas (2010)
6. Korobilis, D., Gilmartin, M.: On regional unemployment: An empirical examination of the determinants of geographical differentials in the UK. Scott. J. Polit. Econ. (2012) doi: 10.1111/j.1467-9485.2011.00575.x
7. Liesenfeld, R., Richard, J.F., Vogler, J.: Likelihood-based inference and prediction in spatiotemporal panel count models for urban crime. J. Appl. Econ. (2016) doi: 10.1002/jae.2534
8. Manolopoulou, I., Chan, C., West, M.: Selection sampling from large data ses for targeted inference in mixture modeling. Bayesian Anal. 3, 429–450 (2010)
9. Paap, R., Franses, P.H., van Dijk, D.: Does Africa grow slower than Asia, Latin America and the Middle East? Evidence from a new data-based classification method. J. Dev. Econ. 77, 553–570 (2005)
10. Ratcliffe, J.H.: Crime mapping: spatial and temporal challenges. Handbook of Quantitative Criminology, Piquero, A.R., Weisburd, D. (eds). Springer: New York, 524 (2013)
11. Roth, R.E., Ross, K.S., Finch, B.G., Luo,W., MacEachren, A.M.: Spatiotemporal crime analysis in U.S. law enforcement agencies: current practices and unmet needs. J. Gov. Inf. 30, 226–240 (2013)
Full Text:
ZIP