Tom Claassen

Assistant Professor - Data Science Institute for Computing and Information Sciences (ICIS) Faculty of Science, Radboud University Nijmegen, Netherlands Tom.Claassen (at) ru.nl / t.claassen (at) science.ru.nl / tomc (at) cs.ru.nl Office: Mercator 1, room 6.03 For more contact details see here

Research

Main topics of interest:

• Statistical Causal Inference (principles and methods)
• Graphical models
• Dynamical systems and time series analysis
• Applications of causal discovery in ecological / (bio)medical / neuropsychological domains

My research focusses on extending principled causal discovery methods to handle the challenges of real-world data and experiments, including: latent confounders, nonlinear interactions, cyclic/feedback mechanisms, missing data, nonstationary systems, and data from multiple, overlapping data sets.

I am currently involved in projects on personalised care in oncology (PersOn,NWO-Perspectief-2023), causal mechanisms in long Covid and respiratory tract infections, sustainable food development and impact on animal movement (SOSFood,EU-Horizon2023), causal mechanisms behind vascular surgery and adverse brain outcomes (AI for Health), night-time dynamics of mother-father-infant triads and impact on wellbeing (Nightly Dance, ZonMw-2023), developing new causal/ML methods to help understand immune cell movements (computational biology), and artificial scientific understanding in ML.

Selected publications

• KG Barman, S Caron, T Claassen, H De Regt. Towards a benchmark for scientific understanding in humans and machines Minds and Machines, 2024 [pdf]
• Tom Claassen, Joris Mooij. Establishing Markov Equivalence in cyclic directed graphs, Uncertainty in Artificial Intelligence (UAI-Best Paper Award), 2023. [pdf]
• Tom Claassen, Gabriel Bucur. Greedy equivalence search in the presence of latent confounders, Uncertainty in Artificial Intelligence (UAI), 2022. [pdf]
• KP Mielke, T Claassen, M Busana, T Heskes M Huijbregts, K Koffijberg, A Schipper. Disentangling drivers of spatial autocorrelation in species distribution models Ecography, 2020. [pdf]
• JM Mooij, T Claassen. Constraint-based causal discovery using partial ancestral graphs in the presence of cycles Uncertainty in Artificial Intelligence (UAI), 2020. [pdf]
• JM Mooij, S Magliacane, T Claassen. Joint causal inference from multiple contexts Journal of machine learning research, 2020. [pdf]
• S Magliacane, T Van Ommen, T Claassen, S Bongers P Versteeg JM Mooij. Domain adaptation by using causal inference to predict invariant conditional distributions Advances in neural information processing systems, 2018. [pdf]
• E Sokolova, AM Oerlemans, NN Rommelse, P Groot C Hartman, JC Glennon, T Claassen, T Heskes, JK Buitelaar. A causal and mediation analysis of the comorbidity between attention deficit hyperactivity disorder (ADHD) and autism spectrum disorder (ASD) Journal of autism and developmental disorders, 2017. [pdf]
• S Magliacane, T Claassen, JM Mooij. Ancestral Causal Inference. Advances in Neural Information Processing Systems (NeurIPS), 2016. [pdf]
• T Claassen, JM Mooij, and T Heskes. Learning Sparse Causal Models is not NP-hard. Uncertainty in Artificial Intelligence (UAI), 2013. [pdf]
• Tom Claassen, Tom Heskes. A Bayesian Approach to Constraint Based Causal Inference Uncertainty in Artificial Intelligence (UAI-Best Paper Award), 2012. [pdf]
• Tom Claassen, Tom Heskes. A logical characterization of constraint-based causal discovery. Uncertainty in Artificial Intelligence (UAI), 2011. [pdf]
• Tom Claassen, Tom Heskes. Causal discovery in multiple models from different experiments. Advances in Neural Information Processing Systems 23, pp.415-423, 2010. [pdf] [bib]