Graph Neural Networks Application in crime prediction - Adaptation of a traffic accident model

• Grant number: 23/04137-2
• Support Opportunities: Scholarships in Brazil - Scientific Initiation
• Start date: July 01, 2023
• End date: June 30, 2024
• Field of knowledge: Physical Sciences and Mathematics - Computer Science - Computing Methodologies and Techniques
• Principal Investigator: Luis Gustavo Nonato
• Grantee: Marvin Mendes Cabral
• Host Institution: Instituto de Ciências Matemáticas e de Computação (ICMC). Universidade de São Paulo (USP). São Carlos , SP, Brazil
• Associated research grant: 22/09091-8 - Criminality, insecurity, and legitimacy: a transdisciplinary approach, AP.ESCIENCE.TEM

Abstract

Undeniably, the crime situation in large urban areas is becoming increasingly serious. The contribution of science with more accurate and effective methods to predict and prevent occurrences that affect the safety of society grows every day. Machine learning and data science have played a key role in this public safety context. With the advent of Deep Neural Networks (DNN), the expectation of more effective methodologies for the analysis and prediction of crimes opens up, making it possible to use a large number of variables in the construction of predictive models. In addition, it becomes feasible to make predictions at a very fine granularity level. This project aims to study and implement a specific type of DNN for analysis and prediction of crimes, the so-called Graph Neural Networks, which make it possible to investigate phenomena at the level of street maps, a fact still little explored in the literature.