Group for data mining and machine learning

The group conducts research on data mining methods and their application to predictive modeling and the discovery of patterns in data. In line with current trends, a significant part of the research focuses on methods applicable to large volumes of structured data represented as graphs (networks), trees, and sequences, which can describe complex interdependencies and hierarchical relationships among the factors underlying the phenomenon being modeled. Application domains include text classification, prediction of biological data, prediction of a patient's future condition based on medical history, automated assessment of student programs, automated selection of algorithms for solving problems, prediction of the citation counts of scientific papers, discovery of patterns and rules in business data, prediction of business trends, and others. Techniques for working with NoSQL databases, commonly referred to as "big data" technologies, are also used for organizing and processing large volumes of data.

Some information about research results in this area is available at the following addresses:

• http://bioinfo.matf.bg.ac.rs/
• http://www.matf.bg.ac.rs/~nikolic/
• http://www.matf.bg.ac.rs/~jgraovac/