Laboratory of Parallel Algorithms
The laboratory focuses on the broadly understood computational aspect of parallel algorithms, with particular emphasis on machine learning algorithms, and the possibility of their effective use on large computing clusters.
Due to the increasing amount of data available every day, traditional machine learning algorithms are becoming insufficient and the serial processing paradigm is computationally inefficient. In order to meet the new challenges related to the growing amount of data, in many cases it is necessary to use large computing clusters, which forces the adaptation of the algorithms used to work in parallel mode. In the Laboratory of Parallel Algorithms we consider both theoretical and practical aspects related to this task. In particular, we focus on the following areas: computer vision, tensor computing, deep networks, low quality image processing, underwater image recognition, hyperspectral data classification, histopathological data classification, data unbalance.
Within the considered domains, we develop new algorithms that use parallel computations, in particular, we study the theoretical and practical aspects related to this phenomenon.
Contact: Bogusław Cyganek, cyganek [at] agh.edu.pl