Current research projects

These are the current research projects. You can also view the list of previous research projects.

FWO WOG - Guiding networked societies, linking data science and modelling

Networks of interconnected autonomous computing entities more and more support our society, interacting and influencing each other in complex and unforeseen ways. Examples are smart grids, intelligent traffic lights, logistics and voluntary peer­to­peer clouds as well as socio­technical systems or more generally the Internet of Things. Understanding the characteristics and dynamics of these systems both at the local and global scale is crucial in order to be able to guide such systems to desirable states.

Opinion Dynamics and Cultural Conflict in European Spaces (ODYCCEUS)

The project aims at developing new empirically grounded theoretical foundations and new interactive and participatory tools for understanding, analyzing, visualizing and monitoring the emergence of social con icts and crises driven by di erences in culture and by diverging world-views, with a focus on their distribution in the European geopolitical and media space.

Learning optimal preventive strategies to mitigate epidemics of latent infectious diseases

Viruses that cause latent infectious diseases (e.g. HIV, hepatitis C, papilloma) pose an important threat to public health. The most efficient way to mitigate epidemics of latent infectious diseases is by prevention. However, the complex dynamics of such epidemics render the development of effective and efficient prevention strategies challenging. To support the decision making with respect to prevention strategies, epidemiological models are frequently used. In this research project we aim to improve the state-of-the art of epidemiological models and to use reinforcement learning to learn optimal preventive strategies in epidemiological models.

Multi-Objective Decision Making with Guarantees

We investigate the possibilities of Multi-objective Decision Making (MODeM), and particulary Reinforcement Learning (MORL) for decision support. Specifically, we are interested in providing guarantees with respect to user utility.

Life-Long Hierarchical Reinforcement Learning

Reinforcement Learning is currently applied to single tasks, for instance parking a car, playing a video game, navigating to a goal, etc. Hierarchical Reinforcement Learning allows to decompose a complex task into simpler sub-tasks. For instance, an agent may progressively learn to grab objects, turn them, open doors, navigate through hallways (with doors), then do something interesting in a complete building. This project applies Hierarchical Reinforcement Learning to very complex tasks, for which the agent has to learn many skills.

Coordinating Human and Agent Behaviour in Collective-Risk Scenarios

Different situations, wherein humans interact among themselves or through technologies in hybrid socio-technical systems, resemble social dilemmas, i.e. situations wherein participants have to select between short-term personal profits and long-term social benefits. The behavioural outcome in those dilemmas is very much dependent on how successful the participants are in calculating the risk associated to the uncertainty of future rewards and on anticipating the opponents’ choices.

Fleet Reinforcement Learning for Wind Farm Control

The world is a connected place, in which the cloud plays an increasingly vital role. One example is the Internet of Things, which has a topic of conversation in virtually all industries. The control of physical devices is no exception: modern wireless sensors allow these devices to move forward from local controllers towards smarter cloud-based architectures.

C-CURE: Cost-Sensitive Dynamic User Authentication with Reinforcement Learning

C-CURE addresses a pressing problem with regards to user authentication: lack of flexibility. Current authentication methods are uniform and static. Their power, and thus the involved overhead for the user, is the same for small (in terms of the required security) and large processes or transactions. C-CURE aims to develop a flexible multi-modal approach which adapts the security level and procedure depending on the expected risk and the context of the user.

SeCloud: Security-driven Engineering of Cloud-based applications

The conception of a holistic & coherent set of tools, technologies and techniques that will allow the software industry to proactively think about security in their Cloud-based applications whether SaaS or Mobile. The four considered perspectives are architecture, infrastructure, programming and process.

ArTificial Language uNdersTanding In robotS (ATLANTIS)

ATLANTIS attempts to understand and model the very first stages in grounded language learning, as we see in children until the age of three: how pointing or other symbolic gestures emerge from the ontogenetic ritualization of instrumental actions, how words are learned very fast in contextualized language games, and how the first grammatical constructions emerge from concrete sentences.

A SCAlable and modular system for eneRGY trading between prosumers (SCANERGY)

SCANERGY will be built up by an intelligent multi-agent system that can manage the electricity produced and consumed both on a lower level (dwellings and neighbourhoods) as well as on a higher level (cities). The SCANERGY system enables the smart energy trade between prosumers, while coping with the inherent real-time dynamism in electricity demand and supply.

Stable MultI-agent LEarnIng for neTworks (SMILE-IT)

SMILE-IT aims to extend and improve upon the state of the art in multi-agent reinforcement learning and network management in order to implement and validate a generic, stable, and robust multi-agent reinforcement learning framework, capable of (semi-)automatically managing modern networked systems (e.g., telecommunications networks, smart grids, air traffic routing, traffic control) through software.


How do we really respond in iterated games – inferring strategies from behavioral experiments in the Prisoners Dilemma game

Cooperators should be evolutionary extinct because cheaters always benefit more than cooperators. However cooperation is all around us and without it human society, could not exist. One of the mechanisms which could explain how cooperation emerged and persisted is direct reciprocity, under which the individuals cooperate with each other to ensure future cooperation. A number of strategies emerge in theory which would be the most beneficial to the players in this setting and also lead to high cooperation. However it is still uncertain if these are the strategies people actually use. 

Here, I will conduct a series of Iterated Prisoner's Dilemma experiments where large numbers of players play long games through a computer interface. Typically experiments which search for strategies have very small number of rounds and the resulting level of cooperation is not very high. However, in longer experiments after the initial drop, the level of cooperation is rising to a very high level. The question is what are the strategies that people use when cooperation is established and what kind of behaviour drives this change. Afterwards, I will perform experiments which will test how outside contexts (like the information the players receive, etc.) influence the strategies. Finally, I will pay special attention to the possibility of negotiation during the game. We will test if the inferred strategies explain the observed behaviours using agent based simulations and evolutionary models. 



Identifying the mechanisms involved in transducing binding information through SH3-SH2 supradomains and their role in regulating the activity of members of the family of Src kinases.

Communication is essential at all biological scales. Our cells have been rigged with communication systems that allow them to respond properly to environmental cues. These systems, or signaling pathways, are composed of different proteins, whose activity is regulated by other proteins. This intricate regulatory system ensures that cells provide the right response, both in kind and magnitude. Errors in regulation often result in diseases.

BRiDGEIRIS - Brussels big data platform for sharing and discovery in clinical Genomics

Genome or Exome screening has started to become a norm in clinical settings, with an aim to provide an improved and effective diagnosis process for patients. To cater to this need, hospitals need an effective, certified and reliable (Bio)informatics solution that can merge, store and analyze the multidimensional data forms. Also, challenge is to handle the huge avalanche of data generated from sequencing. This project has been planned to design and develop solution to afore mentioned needs of Centre for Medical Genetics of VUB UZ Brussel, ULB Erasme and UCL, De Duve.

An integrated Methodology to bring Intelligent Robotic Assistive Devices to the user (MIRAD)

The overall objective of the project is to develop and validate an integrated methodology to design intelligent robotic assistive devices and to test these devices in clinical conditions.

The integration relates to all aspects of such devices and their interaction with humans: engineering aspects (mechatronic design, intelligent control, and simulation), physiological and clinical aspects, and psychological aspects.

Adaptive Preferences for a Changing World: an AI study in the co-evolution of preferences and group formation.

Many strategic and economic situations are characterized by participant preferences that take into account the well being of others.  These preferences guide humans in their choices with whom to participate in economical or social activities.  Little attention has been given to how this group formation shapes these preferences and, vice versa, how these preferences shape formation of groups.

Towards an understanding of the interplay between other regarding preferences and group formation in strategic environments

Many strategic situations are characterized by player’s preferences that take into account the well being of others.  These preferences guide humans in their choices with whom to participate in economical or social activities.  Little attention has been given to how group formation shapes players’ beliefs concerning preferences and how preferences guide the formation of groups.

Multi-agent reinforcement learning of coordination and problem structure

In this project, we will investigate the automatic detection of interactions between agents and its use for local coordination, without requiring agents to have a full view of other agent’s state and actions. Furthermore, we will validate the techniques developed from gained insights on settings that are fully cooperative, fully competitive and mixed.

Personalized Products Emerging from Tailored User Adapting Logic (Perpetual)

The ambition of the PERPETUAL project is to develop methods and algorithms that achieve automatic, user-oriented functional demand profiling in support of enhancing product service performance, while requiring less resources through an increased efficiency as a result of well-optimized control policies for a wide range of applications.