Johan Kwisthout, hoogleraar Probabilistic Graphical Models
Johan Kwisthout, hoogleraar Probabilistic Graphical Models

Johan Kwisthout appointed professor of Probabilistic Graphical Models

From 1 January, Johan Kwisthout has been appointed professor of Probabilistic Graphical Models at Radboud University's Faculty of Social Sciences.

Kwisthout researches how AI techniques such as Bayesian networks (a technique for dealing with uncertain information) can responsibly help patients and healthcare providers make decisions about treatments and follow-up care in cancer. 'AI assists in calculating possible scenarios, comparing alternatives, and explaining advice,' Kwisthout explains. 'Specifically, I am interested in transparency and explainability: how to ensure that the AI can motivate its own reasoning and, above all, explain its own uncertainty. It's crucial that the AI not only looks at purely medical aspects, but can also include quality of life in the advice.'  

Kwisthout: 'The field of AI faces important questions. How can we use technology sustainably and responsibly to support people, rather than irresponsibly dealing with privacy, data, energy, and human labour? Sustainable, transparent, and reliable AI must be at the forefront. That is what I am going to fight for!' 

About Johan Kwisthout 

Kwisthout (Breda, 1976) graduated in Technical Computer Science from the Open University in 2005, and in Artificial Intelligence from Radboud University in 2006. He then obtained his PhD at Utrecht University on a study of the complexity of different types of computation on Bayesian networks.  

After several temporary appointments as lecturer and researcher in Nijmegen and Leiden, he was appointed associate professor of Artificial Intelligence in Nijmegen in 2016. He has been associate professor since 2020, and is now also programme director of the Artificial Intelligence programme.