UCL Department of Geography


Description Photo Here

Personal tools
Log in
This is SunRain Plone Theme
UCL Home  /  Geography  /  People  /  Research Students  /  Victor-Alexandru Darvariu

Victor-Alexandru Darvariu


victor-alexandru_darvariu-sml.jpgPhD Student

North-West Wing
UCL Department of Geography
Gower Street
London WC1E 6BT

+44 (0) 20 3108 4413 x54413

Personal website


PhD Title: Building and Enhancing Networked Systems Through Multi-Agent Deep Reinforcement Learning

Abstract: Many systems can be conceptually thought of as networks: from the complex web of social interactions humans engage in every day to much of the infrastructure that powers our society. Victor is interested in studying the properties of these networks, with a particular focus on how behaviours or patterns emerge and evolve with time. Aside from developing a theoretical framework for reasoning about these systems, he aims to examine its applications in areas such as communication or transport networks. To realise this, he draws on techniques from the fields of network science, machine learning (with a particular focus on reinforcement learning), and artificial intelligence.

Supervisors: Professor Mirco Musolesi and Professor Steve Hailes.


Victor-Alexandru Darvariu is a second year doctoral student at University College London and The Alan Turing Institute. Before starting his graduate studies, he gained his MSci Computer Science with an Industrial Year degree at the University of Birmingham. Victor previously held Research Assistant posts at both UCL and Birmingham, working in several projects in the data science realm. He also has experience as a software engineer on a variety of platforms, having built everything from mobile apps to computer security software in the financial sector.


  • A Functional Perspective on Machine Learning via Programmable Induction and Abduction. Steven W. T. Cheung, Victor-Alexandru Darvariu, Dan R. Ghica, Koko Muroya and Reuben N. S. Rowe. In Proceedings of the Functional and Logic Programming Symposium (FLOPS) 2018, Nagoya, Japan.