Uncertain Reasoning, Data Mining and Big Data
How can we teach a computer to reason and take decisions in uncertain situations? For example, we might want to teach a computer to assist a doctor in medical diagnosis; or to support environmental experts in the prediction of natural hazards, such as landslides; or to help recognizing potential terroristic threats carried out over the internet, or by airplanes. We can do this by transferring human expertise into a knowledge base and developing algorithms for logical reasoning under uncertainty that are applied to the knowledge base to perform inferences, such as deductions or predictions.

In some cases human expertise is not available but there are data. It can be data in the medical domain (such as clinical information about patients, or genetic data), in business (e.g., credit cards, insurance), or environmental data, just to say a few. In these cases, it is important to give a computer tools to be able to automatically learn about the domain from the available data.

Data mining is the field that enables computers to learn from data how to do predictions, diagnosis, recognition (and more), using very little human intervention. Applications include again medical diagnosis, predictions, user profiling, recommendations, image/speech recognitions.

Our expertise includes many applications of the above methodologies to real world problems, such as the automated diagnosis of dementia (medicine), the prediction of debris flows (environment), and the identification of flying objects in a no-fly area (defense). At the same time we have a long experience in the field of biomedical modelling and in particular, genomics.


  • Business Engineering and Production
  • Energy and Environment
  • Industrial Technologies
  • Information and Communication Technologies

For an overview of some of our applied projects please have a look to our lealeft of Imprecise Probability Group