Data science

Data science can be regarded as the "fourth paradigm" of science (empirical, theoretical, computational and now data-driven). The researchers of the Institute of Mathematics also take part in the research of this emerging interdisciplinary scientific field.

Our work focuses on three main areas:

Research and development in cooperation with our industrial partners. We have been participating in many successful data intensive R&D projects with renowned companies throughout the years, for more details see the homepage of the Statistics and Mathematical Modeling Consulting Group. One of our most important partners is Nokia Bell Labs with whom we have long-lasting and fruitful cooperation that is also reported by the international webpage of Nokia Bell Labs.

- Probabilistic graphical model based machine learning. Probabilistic graphical models are fundamental in understanding and building intelligent systems. Their graph theoretic properties provide both an intuitive interface to model the interacting factors, and a data structure facilitating efficient learning and inference.  Our research focuses on developing the theory and algorithms for discovering and handling the information hidden in the data. We also apply the new methodologies to real problems raised in our cooperation with industrial partners.

- Data-driven educational research. Machine learning approaches are also emerging in social sciences e.g., in education research.  This research is carried out in cooperation with the Central Academic Office. Using statistical learning techniques, we predict university dropping-out based on secondary school performance, we study the predictive power of admission point score on university performance, we investigate the effects of remediation courses, and we study the impact of grade inflation on student evaluation of teaching. We also develop tools to gather, visualize and analyze educational data to help policymakers and students as well.

Our work is also recognized by funding agencies. We are supported by

  • the EFOP-3.6.2-16-2017-00015 project entitled ”Deepening the activities of HU-MATHS-IN, the Hungarian Service Network for Mathematics in Industry and Innovations and by
  • the Higher Education Excellence Program of the Ministry of Human Capacities in the frame of Artificial Intelligence research area of Budapest University of Technology an Economics (BME FIKP-MI/SC).