Description
Leader of the group Machine Learning: Peter Grünwald.
Our research group focuses on how computer programs can learn from and understand data, and then make useful predictions based on it. These algorithms integrate insights from various fields, including statistics, artificial intelligence and neuroscience.
Machine-learning applications are increasingly part of every aspect of life, from speech recognition on cell phones to illness prediction in healthcare. One common problem is extremely polluted data, for which no single model can provide adequate explanations. At CWI we address this issue with statistical machine learning based on combining predictions from different models and experts in order to achieve reliable conclusions.
We also study how networks of neurons in the brain process information, and how modern deep-learning methods can benefit from neuroscience. We develop novel neural networks, like Deep Adaptive Spiking Neural Networks, and also theoretical models of neural learning and information processing in biology. Applications of our work range from low-energy consumption neural machine learning to neuroprosthetics, to increased insight into the question of how the brain works.
Vacancies
Post-doc position in the area of Spiking Neural Networks
The Machine Learning group at the Centrum Wiskunde & Informatica (CWI) in Amsterdam invites applicants for a Post-doc position (2 or 3 years) in the area of Spiking Neural Networks. We are looking for a motivated postdoctoral researcher with expertise in the area of spiking neural networks and an interest in probabilistic computing.
News
Energy-efficient AI detects heart defects
Mathematical breakthrough by Centrum Wiskunde & Informatica (CWI) allows neuromorphic chip to recognise heart defects, speech and hand gestures up to a thousand times more energy efficient. Such low-energy AI chips are ideal for wearable and mobile applications.
CWI launches first podcast ‘Superscience’
Together with IT professional magazine AG Connect, CWI presents a podcast series in honour of its 75th anniversary: Superscience. In the first episode: Sander Bohté.
Jan Hemelrijk award for CWI's Rosanne Turner
The Netherlands Society for Statistics and Operations Research (VVSOR) awarded the Jan Hemelrijk Award to PhD student Machine Learning Rosanne Turner.
Bayesian learning from data: challenges, limitations and pragmatics
How do humans and computers learn from data? PhD student Rianne de Heide of CWI’s Machine Learning group explored Bayesian learning in specific.
Members
Associated Members
Publications
-
Hendriksen, A.A, de Heide, R, & Grünwald, P.D. (2021). Optional stopping with Bayes factors: A categorization and extension of folklore results, with an application to invariant situations. Bayesian Analysis, 16(3), 961–989. doi:10.1214/20-BA1234
-
Dora, S, Bohte, S.M, & Pennartz, C. (2021). Deep gated Hebbian predictive coding accounts for emergence of complex neural response properties along the visual cortical hierarchy. Frontiers in Computational Neuroscience, 15. doi:10.3389/fncom.2021.666131
-
de Graaf, H, ter Schure, J.A, & van Liere, G.A.F.S. (2021). How old are young people when they start having sex? Unravelling the applicability of Cox proportional hazards regression. Journal of Public Health. doi:10.1007/s10389-021-01619-6
-
Mücke, N.T, Bohte, S.M, & Oosterlee, C.W. (2021). Reduced order modeling for parameterized time-dependent PDEs using spatially and memory aware deep learning. Journal of Computational Science, 53. doi:10.1016/j.jocs.2021.101408
-
Zambrano, D, Roelfsema, P.R, & Bohte, S.M. (2021). Learning continuous-time working memory tasks with on-policy neural reinforcement learning. Neurocomputing. doi:10.1016/j.neucom.2020.11.072
-
de Heide, R. (2021, January 26). Bayesian Learning: Challenges, Limitations and Pragmatics.
-
Proença, H.M, Grünwald, P.D, Bäck, T, & van Leeuwen, M. (2021). Discovering outstanding subgroup lists for numeric targets using MDL. In ECML PKDD: Joint European Conference on Machine Learning and Knowledge Discovery in Databases (pp. 19–35). doi:10.1007/978-3-030-67658-2_2
-
de Heide, R, Kirichenko, A.A, Mehta, N.A, & Grünwald, P.D. (2020). Safe-Bayesian Generalized Linear Regression. In Proceedings of the International Conference on Artificial Intelligence and Statistics (pp. 2623–2633).
-
Pozzi, I, Bohte, S.M, & Roelfsema, P.R. (2020). Attention-Gated Brain Propagation: How the brain can implement reward-based error backpropagation. Advances in Neural Information Processing Systems.
-
Turner, R.J, & Grünwald, P.D. (2020). Safe tests for 2 x 2 contingency tables and the Cochran-Mantel-Haenszel test.
Software
Squint: Experimenting in Prediction with Expert Advice problems
Squint provides a codebase to perform numerical proof-of-concept experiments in learning theory, particularly in Prediction with Expert Advice problems, a core problem in learning theory.
Current projects with external funding
-
Efficient Deep Learning Platforms (eDLP)
-
Enabling Personalized Interventions (EPI)
-
Human Brain Project - SGA3 (HBP-SGA3)
-
Bayesian Best-Arm Identification: optimal algortihms for sequential decision making (None)
-
Efficient Models of Decision-Making for Asseing Cognitive Processing States (None)
-
Perceptive acting under uncertainty: safety solutions for autonomous systems (None)
-
Safe Bayesian Inference: A Theory of Misspecification based on Statistical Learning (SAFEBAYES)
-
Spiking Neural Networks research program
Related partners
-
INRIA
-
Katholieke Universiteit Nijmegen
-
Philips
-
KPMG
-
SURFsara B.V.
-
Technische Universiteit Eindhoven
-
Technische Universiteit Delft
-
Universiteit Twente
-
Universiteit van Amsterdam
-
Vrije Universiteit Amsterdam