Amirhossein Ghaffari

I am a doctoral researcher in Computer Science and Engineering at the Center for Ubiquitous Computing, Faculty of Information Technology and Electrical Engineering (ITEE), University of Oulu, Finland. My work lies at the intersection of urban well-being, machine learning, and graph neural networks, with a focus on developing Graph Neural Networks, and in particular Spatio-Temporal Graph Neural Networks (STGNNs), for city-scale predictions and intelligent decision support. Combining ideas from machine learning, network science, and urban analytics, I aim to model dynamic, data-driven systems that enhance urban well-being and sustainability.

Alongside my doctoral work, I have contributed to projects spanning federated learning, smart city optimization, and machine learning for environmental and communication systems. My background includes computer vision, signal analysis, and predictive modeling using both deep learning and classical ML approaches.

I am passionate about turning heterogeneous data, from sensors, networks, and open urban datasets, into actionable insights through scalable, interpretable, and privacy-preserving AI models.

Before starting my PhD, I worked for several years as a computer vision and deep learning researcher at Fard Iran Co., designing algorithms for automatic number plate recognition, weigh-in-motion (WIM) systems, and intelligent surveillance, and helping deploy these systems at scale. I hold an M.Sc. from Sharif University of Technology and a B.Sc. from Amirkabir University of Technology.

My current work spans spatio-temporal graph learning, traffic and mobility modeling, and broader questions around how heterogeneous urban data can be processed and used efficiently. I am always happy to discuss research, collaboration, or interesting opportunities feel free to reach out by email or connect on LinkedIn.