Objective
Exploiting call detail record (CDR) data provides deep and granular insights into the interactions and behaviors of the population in a city. These data, which record information on the use of telecommunications networks, are a valuable source for understanding patterns of mobility, population density and economic activity.
This project proposes a methodology for the identification of regions in the city of Madrid using Graph Embedding techniques from CDR data. The main objective is to determine significant clusters within the city that can be exploited through stochastic models and descriptive studies. To achieve this, a graph is constructed where the nodes represent communication towers (antennas) and the edges reflect the intensity of connections made by foreign users, i.e., the trips made by tourists are counted based on the sequence of antennas to which they have connected over time.
Advanced Graph Embedding techniques are applied to convert the graph structure into a low-dimensional vector space, preserving the essential topological properties. These embeddings allow the identification of communities and communication behavior patterns. The results obtained provide a solid basis for the analysis of demographic, socioeconomic and mobility characteristics within the different regions, facilitating the implementation of predictive models and the development of informed urban policies. This innovative approach not only improves the understanding of Madrid’s urban dynamics, but also opens up new possibilities for the efficient planning and management of urban resources.
PABLO ZUBASTI RECALDE
Degree
Computer Science Engineering, Universidad Carlos III de Madrid (September2018 – June 2024)
Work experience:
Researcher at Universidad Carlos III de Madrid. Projects: SIMBAT project: Solutions for Intelligent Monitoring based on drone data and AI Tools (PDC2021-121567-C22) y Modelos de datos para fenómenos Precipitación-Escorrentía-Data-driven models for rainfall-streamflow events (TED2021-131520B-C22).
Intern at Cátedras Grupo MásMóvil (MásOrange).
Technical skills:
Programming languages: C/C++, Python, R, Java, Matlab JavaScript and ARM.
Tools: WSL, Docker, Visual Studio Code, Vim, Git, Eclipse, Pycharm, Unreal Engine, Unity, Blender, QGIS and ArcGIS.
Databases: Oracle, MySQL, MariaDB and Big Query.
Extra: GCP and LaTeX.