Pavel Kromer


Pavel Kromer

VSB-TU Ostrava, Czech Republic

Research field: Evolutionary algorithms, data mining, gpgpu computing.

Biography and Abstract


Pavel Krömer is an associate professor of Computer Science. He obtained M.Sc. and Ph.D. in Computer Science from the VSB – Technical University of Ostrava, Czech Republic, in 2006 and 2010, respectively. Between 2005 and 2010, he was employed as a software specialist in a large software company. In 2010, he returned to academia as an assistant professor and junior researcher at the Department of Computer Science, VSB – TU Ostrava. During 2014, he served as a postdoctoral fellow at the Department of Electrical and Computer Engineering, University of Alberta, Edmonton, Canada. Upon returning to the VSB – TU Ostrava, he gained the rank of associate professor and senior researcher. He is now affiliated with the VSB – TU Ostrava and the IT4Innovations National Supercomputing Center, where he assumes a role of External Specialist. He is a member of the IEEE, System, Man, and Cybernetics Society and serves on SMC’s Technical Committees on Soft Computing and Big Data Computing. He is also a member of steering committee of the Neural Network World Journal and served in various chairing roles at several international conferences including IBICA 2014 (publicity co-chair), AECIA 2015, 2016 (publicity co-chair), IBICA 2015 (PC co-chair), INCoS 2016, 2017 (workshop co-chair, international liaison co-chair), etc.

Pavel’s areas of interest include computational intelligence, information retrieval, data mining, knowledge discovery, and parallel and distributed computing. He held courses on information retrieval and programming and nowadays provides university courses on parallel and distributed computing.



Talk: Hybrid Intelligent Data Mining: An example of one multi-paradigm approach

Abstract: Intelligent methods (metaheuristics) form a class of algorithms that try to apply various unconventional approaches to efficiently solve problems. Nature–inspired metaheuristics mimick successful optimization strategies from nature. The strategies are applied to iteratively develop single or multiple problem solutions and aim at adaptability and reusability. They include methods of swarm intelligence, evolutionary computation, and e.g. neural computation and are considered a part of the wide family of Computational Intelligence methods. Hybrid intelligent methods is an umbrella term referring to algorithms combining two or more intelligent approaches into a coherent framework addressing some theoretical or practical problem. This talk provides an overview of selected intelligent methods, describes the anatomy of one hybrid algorithm, and points out some of its applications.