Euclid is one way to see the world. F-B, U-D, L-R, in a binary fashion. What about Power, Revolution, Language, History? Orientation works differently in non-euclidean space: it is biased and political. Networks, represented as graphs, have manifold geometries with spectral metrics. Social networks, transportation systems, infrastructures, landscapes, logistics, epidemics, …, it is all about signals. Exchanging information.

A node can be identified through its interactions with other nodes in the network. An exciting approach to this topic is representation learning on graphs as exemplified by modern researchers who merge graph theory, signal processing and machine learning.

My research focus lies in cities as complex systems, geospatial analytics, spatiotemporal dynamics and infrastructures. Here I share my thoughts on technical matters. I provide data analytic consultancy services for fun and profit. Looking for opportunities in academia or industry.