Our everyday interactions shape the spirit of the city we live in. That is especially true for the -great- cities, London, New York, Berlin, which exhibit complex and changing spatio-temporal properties. We implicitly learn some of these properties over time. Seasoned New Yorkers, that are interested in their safety, have learned to avoid certain streets on their way to work, whereas young inhabitants of Berlin tend to cluster in the neighborhood of artists in their nights out. But what other secrets do great cities hide?

The objective of this project is to uncover, understand and characterize the hidden aspects of the inner life of cities. We will use two key ingredients to do so. The first is a plethora of data [1,2]. Fortunately, many cities have realized the importance of open data for their operation and there have undergone a systematic effort in exposing their inner workings. New York for example recently released a huge dataset listing the entirety of criminal incidents within its borders [3]. This dataset can potentially uncover the shift in the inner workings of the criminal underground. The second ingredient of this project is harmonic analysis over graphs [4,5]. Recent advances in signal processing and machine learning have provided tools that enable the modeling the complex relationships present in many high dimensional datasets. These tools are ideal for cities because they can deal concurrently with the spatial and temporal aspect of the data.

Interested? Contact Andreas Loukas (andreas.loukas@epfl.ch) for more details.

[1] https://data.cdrc.ac.uk [2] https://data.cityofnewyork.us/ [3] https://data.cityofnewyork.us/Social-Services/311-Service-Requests-from-2010-to-Present/erm2-nwe9 [4] The Emerging Field of Signal Processing on Graphs, David I Shuman, Sunil K. Narang, Pascal Frossard, Antonio Ortega, and Pierre Vandergheynst, 2013. [5] Frequency Analysis of Temporal Graph Signals, Andreas Loukas, Damien Foucard, 2016.