Modelling crime and its fear

Day - Time: 20 April 2017, h.14:00
Place: Area della Ricerca CNR di Pisa - Room: C-29
  • Rafael Prieto Curiel (University College London)

Fosca Giannotti


Generally speaking, crime is a rare event, which means that traditional tools for modelling, simulating or even measuring crime are not appropriate because they are not comparable across time and populations. Its low frequency also does not correspond to the more frequent fear of crime, which means that often people are more fearful than victimised.
An estimate of the distribution of crime suffered by the population based on a mixture model will be presented, which allows new and standardised measurement of the concentration of rare events, allowing us to determine if a policy results in victim displacement rather than crime reduction, and an example using victimisation data from Mexico City will be presented.
This modelling technique allows us to consider the quantify and simulate the contagion process of the fear of crime which explains the observed miss-match between crime and its fear. Similar models might also be used for other rare events, such as human mobility patterns

Rafael Prieto Curiel: I am an applied mathematician from Mexico. I worked for the Police Department in Mexico City with spatial models of crime and policing. I did an MSc in Statistics at University College London, where now I am doing a PhD in Mathematics. I started a Mathematics Magazine called Chalkdust and I am currently its Editorial Director.