Un indicador de criminalidad procesada en Colombia mediante análisis de componentes principales ponderados geográfica y temporalmente

Crime analysis is essential to put in place efective security policies. This study focuses on creating an index to assess borough prosecuted crime in Colombia from 2018 to 2023. The index integrates variables of criminal offences along with population density to identify spatiotemporal patterns of criminality and high-risk hotspots through geographically and temporally weighted principal component analysis. The implementation of a robust tool allowed a better explanation of the criminal phenomenon, highlighting the relevance of autocorrelation and spatial-temporal heterogeneity in its analysis. A comparison of spatial multivariate analysis methodologies was carried out by evaluating different kernel functions and bandwidths to optimise the models. The results signal a strong spatial heterogeneity in crime, which suggest the presence of non-random patterns determined by geographical and temporal elements. These findings provide a more precise understanding of the spatial distribution of crime. Finally, a spatiotemporal indicator is created to allow for analysis and comparison across regions and time periods, facilitating the implementation of more effective preventive and corrective measures in the most affected areas.