Welcome to the Algorithmics group, a section of the Department of Software Technology of the Faculty of Electrical Engineering, Mathematics and Computer Science at Delft University of Technology.
In the Algorithmics group, we aim to design, and understand fundamental properties of, planning and coordination algorithms for intelligent decision making in real world applications, such as for coordinating electrical loads within network constraints, or logistic processes on a shunting yard or container terminal. Apart from dealing with the inherent scalability challenge in such combinatorial domains, we work on two main research objectives. First, we aim to understand how to build predictive models from data that facilitate effective planning and scheduling algorithms dealing with uncertainty, for example to find robust solutions. Our second research objective is to understand how to effectively incorporate behaviour and preferences of stakeholders in the decision-making process.
To realize these objectives we focus on the following research topics:
- Model-based reinforcement learning: How to deal with uncertainty in planning?
- Multi-party optimization: How to account for preferences and possibly strategic behaviour in multi‐actor planning?
- Constraint reasoning: How to support individual and group scheduling decisions?
- Meta-heuristics: What are effective general heuristic methods to support decisions taking into account different preferences?
We aim to apply these techniques in sectors such as logistics and transportation, energy, and maintenance. To realise these applications we pay attention to handling large amounts of available data as well as to the development of advanced algorithmic toolboxes based on generic methods and techniques.
- Dr Matthijs Spaan and Dr Neil Yorke-Smith have two openings for PhD positions on uncertainty in AI. The positions are part of the FET Open project Epistemic AI. The aim is to develop novel reinforcement learning and constraint programming algorithms under non-probailistic uncertainty. Deadline for applying is the end of February 2021.
- Mathijs de Weerdt and Matthijs Spaan have two openings for PhD positions in the AI4B.io project: one on smart plant scheduling with Dr. De Weerdt and one on reinforcement learning for a self-driving lab with Dr. Spaan. Deadline for application is March 15, 2021.
- Our group holds a number of openings for MSc projects for TU Delft students.
- If you're among the best students in your class (CS/AI/Maths) and interested in obtaining your PhD by contributing to our mission (see above), please send an email to the professor you'd like to be your promotor indicating your topic/direction of interest.
Research and Education for Master Students
28 maart 2018
Neil Yorke-Smith gave an invited talk at the STEP Conference 2018
Neil Yorke-Smith gave an invited talk at the STEP Conference 2018 on Citizenship in the Age of AI (news post at DDFV).
28 maart 2018
Simon Mountakis defended his PhD thesis
Simon Mountakis defended his PhD thesis.
19 maart 2018
Mathijs de Weerdt won the best poster award at ICT.OPEN 2018
Mathijs de Weerdt won the best poster award at ICT.OPEN 2018 with his work on Intention-Aware Routing for Electric Vehicles.
09 augustus 2017
PhD student Erwin Walraven released a software toolbox
PhD student Erwin Walraven released a software toolbox containing accelerated algorithms for solving Partially Observable Markov Decision Processes.
17 maart 2017
PhD student Rens Philipsen was interviewed
PhD student Rens Philipsen was interviewed by the energiepodium.nl
08 december 2016
PhD student Gleb Polevoy has become doctor of the week
PhD student Gleb Polevoy has become doctor of the week (defence December 6, 2016)
01 september 2016
We are organizing the International Conference on Automated Planning and Scheduling (ICAPS) 2018 in Delft
We are organizing the International Conference on Automated Planning and Scheduling (ICAPS) 2018 and the International Conference on Integration of AI and...
18 mei 2016
Erasmus Energy Forum Science Award 2016
Rens Philipsen, German Morales-Espana, Mathijs de Weerdt, and Laurens de Vries won the Erasmus Energy Forum Science Award 2016 with their paper on Imperfect...
01 maart 2016
Proposal granted for Optimising Flexible Energy Use in Industry for the NWO ESIpose program
Our proposal for Optimising Flexible Energy Use in Industry for the NWO ESIpose program was granted (2 PhD students)
08 juni 2015
Mathijs de Weerdt was been elected as teacher of the year 2015
Mathijs de Weerdt was been elected as teacher of the year 2015 of the Computer Science department at TU Delft.
Mathematics & Computer Science (EE)MCS
Building 28 - Van Mourik Broekmanweg 6