A latent Markov model for detecting patterns of criminal activity
The paper investigates the problem of determining patterns of criminal behaviour from official criminal histories, concentrating on the variety and type of offending convictions. The analysis is carried out on the basis of a multivariate latent Markov model which allows for discrete covariates affec...
| Main Author: | |
|---|---|
| Contributors: | ; |
| Format: | Electronic Article |
| Language: | English |
| Published: |
2007
|
| In: |
Journal of the Royal Statistical Society. Series A, Statistics in society
Year: 2007, Volume: 170, Issue: 1, Pages: 115-132 |
| Online Access: |
Volltext (Verlag) |
| Journals Online & Print: | Drawer...
|
| Check availability: | HBZ Gateway |
| Keywords: |
| Summary: | The paper investigates the problem of determining patterns of criminal behaviour from official criminal histories, concentrating on the variety and type of offending convictions. The analysis is carried out on the basis of a multivariate latent Markov model which allows for discrete covariates affecting the initial and the transition probabilities of the latent process. We also show some simplifications which reduce the number of parameters substantially; we include a Rasch-like parameterization of the conditional distribution of the response variables given the latent process and a constraint of partial homogeneity of the latent Markov chain. For the maximum likelihood estimation of the model we outline an EM algorithm based on recursions known in the hidden Markov literature, which make the estimation feasible also when the number of time occasions is large. Through this model, we analyse the conviction histories of a cohort of offenders who were born in England and Wales in 1953. The final model identifies five latent classes and specifies common transition probabilities for males and females between 5-year age periods, but with different initial probabilities. |
|---|---|
| ISSN: | 0964-1998 |