Predicting crime in Bogota using Kernel warping
Predicting crime is fundamental for an efficient use of the currently allocated police force resources. Using a kernel warping methodology developed by Zhou and Mattenson (2016) to predict ambulance demand in Melbourne, we predict crime in Bogota, Colombia. Exploiting the geometry of the city, this...
- Autores:
-
Garrido Mejía, Sergio Hernán
- Tipo de recurso:
- Fecha de publicación:
- 2018
- Institución:
- Universidad de los Andes
- Repositorio:
- Séneca: repositorio Uniandes
- Idioma:
- eng
- OAI Identifier:
- oai:repositorio.uniandes.edu.co:1992/34610
- Acceso en línea:
- http://hdl.handle.net/1992/34610
- Palabra clave:
- Crimen - Investigaciones - Bogotá (Colombia)
Crimen - Predicciones - Bogotá (Colombia)
Economía
- Rights
- openAccess
- License
- http://creativecommons.org/licenses/by-nc-nd/4.0/
| Summary: | Predicting crime is fundamental for an efficient use of the currently allocated police force resources. Using a kernel warping methodology developed by Zhou and Mattenson (2016) to predict ambulance demand in Melbourne, we predict crime in Bogota, Colombia. Exploiting the geometry of the city, this methodology performs worse than the state of the art Kernel Density Estimation (KDE) methodology by a 100 basis points using the Area Under the Curve (AUC) of the Cumulative Accuracy Profile (CAP) metric. The results of this model are robust to changes in the training and testing period and different definitions of the adjacency and Laplacian matrices. |
|---|