Pavlos Andreadis

Pavlos AndreadisPavlos Andreadis is a PhD student in Informatics at the University of Edinburgh and a research assistant for the SmartSociety project. As part of these duties he has built and maintains this website, while also lending a hand with other WP10 administrative tasks.

Graduating from the Technical University of Crete in 2008 as a Production and Management Engineer, Pavlos continued his studies in Chania, obtaining his MSc in Production Systems in 2010. Through these years Pavlos designed, and helped create, a Multi-Agent system for reaching consensus in Multi-Criteria optimization, published a thesis on an approximate MDP solver utilizing maximum-likelihood trajectories, and gained a rich background in Operations Research and Mathematical Optimization.

The following years found Pavlos working part-time as a private tutor (Statistics, Queue Theory, Optimal Control etc.) and programmer, developing a series of Genetic Algorithms and serving in the Greek Army. In the end of the summer of 2013 he joined the SmartSociety project here in Edinburgh.

Pavlos’s interests lie in the fields of Reinforcement Learning, Multi-Agent Systems and, recently, Social Computation and Preference Learning. Together with Subramanian Ramamoorthy (his  first supervisor and his leader in the Robust Autonomy and Decisions (RAD) group) he focuses on Incentives (WP5) and Social Orchestration (WP6). Currently Pavlos is researching Coarse Preferences, or preferences given Coarse Beliefs, while also developing a Recommender System for the SmartShare application, with plans of incorporating it into a social orchestration model.

Pavlos Andreadis is also part of the Agents Research Group (led by his second supervisor Michael Rovatsos)  through which he has presented work on combining structured preferences with Reinforcement Learning.