Project in collaboration with FGV - http://www.visualdslab.com/projects/CrimeAnalytics and UCSP - https://cs.ucsp.edu.pe/ and ICMC https://www.icmc.usp.br/
Visual Data Crime is a research project that aims at creating computational systems to help identify, understand, and predict criminality. The goal is to explore data from space-temporal crime occurrences, safety perception, socioeconomic traits, and amenity features to help specialists and decision-makers deal with criminality. Knowing the wide range of analyses on criminal behavior and its relations with other variables, this project relies on jointly using machine learning, visualization systems, and optimization methods. Machine learning is used to automatically find crime patterns. Visualization frameworks to help decision-makers to understand found patterns as well as directly search for those patterns. Optimization methods are used to tune and help explaining learning machines, and to automatically improve the visualization tools.
Researches
- Germain García Zanabria
- Erick Gomez Nieto, UCSP
- Jorge Poco, EMAp/FGV
- Luis Gustavo Nonato, USP
Project’s Links
Germain Garcia-Zanabria
Computer Scientist | Data Scientist | Researcher
My areas of interest are data visualization, visual analytics, machine learning, data science, crime analysis, crime prediction, dropout analysis, geo-referenced data, Spatiotemporal analysis, and computer science for social goods.
