Achievements

Our prizes, grants and other accomplishments.

Browse

for Students

See what we can offer you, as a Bachelor, Master of PhD level student.

Learn

Research

Interested in our research? Please come in.

Research

Seminar

Come to our weekly meeting.

Discuss

Team

Take a quick tour through our team members.

Meet

Research

Main research direction

We focus on various machine learning models and algorithms, with particular interest in hybrid models, connecting concepts from both classification and clustering methods.

Events

every thursday, 14:00, room 1146

Group's seminar

Classification methods

We are investigating new models able to perform classification. Including hierarchical, support vector based and entropy based models.

Clustering methods

Our group develops new clustering algorithms, based mostly on information and compression theory.

Deep learning

GMUM is also actively working on deep learning paradigm (in the Hinton's sense), as we strongly believe that this is the future of machine learning.

Recent papers

Marek Śmieja, Dawid Warszycki, Average Information Content Maximization - a new approach for fingerprint hybridization and reduction , PLOS ONE vol. 11/1 (2016), e0146666
Krzysztof Misztal, Michał Podgórski, Przemysław Spurek, Jacek Tabor, Arkadiusz Tomczyk , Detection of Elongated Structures with Hierarchical Active Partitions and CEC-Based Image Representation , LECTURE NOTES IN COMPUTER SCIENCE vol. 403 (2016), 159-168
Edward Delp, Jarosław Duda, Neeraj Gadgil, Paweł Korus, Khalid Tahboub, Image-Like 2D Barcodes Using Generalizations of the Kuznetsov–Tsybakov Problem , IEEE T INF FOREN SEC vol. volume 11 issue 4 (2016), 691-703
Mateusz Babiuch, Marek Skomorowski, Bartosz Zieliński, RoughCut-New Approach to Segment High-Resolution Images , Lecture Notes in Artificial Intelligence vol. 9693 (2016), 591-601
Wojciech Czarnecki, Stanisław Jastrzębski, Damian Leśniak, Learning to SMILE(S) , International Conference on Learning Representations 2016 workshop (2016)
Mateusz Juda, Markus Seidl, Matthias Zeppelzauer, Bartosz Zieliński, Topological Descriptors for 3D Surface Analysis , Lecture Notes in Computer Science vol. 9667 (2016), 77-87
Wojciech Czarnecki, Jacek Tabor, Online Extreme Entropy Machines for Streams Classification and Active Learning , Advances in Intelligent Systems and Computing vol. 403 (2016), 371-381
Andrzej Bojarski, Wojciech Czarnecki, Sabina Podlewska, Robust optimization of SVM hyperparameters in the classification of bioactive compounds , J CHEMINFORMATICS vol. 7 (38) (2015), 15
Andrzej Bojarski, Wojciech Czarnecki, Sabina Podlewska, Extremely Randomized Machine Learning Methods for Compound Activity Prediction , Molecules vol. 20 (11) (2015), 20107-20117
Andrzej Bojarski, Wojciech Czarnecki, Sabina Smusz, Dawid Warszycki, Exploiting uncertainty measures in compounds activity prediction using support vector machines , BIOORG MED CHEM LETT vol. 25 (2015), 100-105
more

Achievements

Grants

Theory of missing data

prof. Jacek Tabor, dr Łukasz Struski, dr Marek Śmieja, dr Bartosz Zieliński
2016-2019, NCN OPUS 10

Clustering algorithm, which uses generalized Gaussian distribution and non-normal distributions

dr Przemysław Spurek
2016-2019, NCN SONATA 10

Detectors and descriptors of the key points based on the topological information

dr Bartosz Zieliński, dr Mateusz Juda
2016-2019, NCN SONATA 10

Algorithmical aspects of synchronization

dr Adam Roman
2016-2019, NCN OPUS 9

Application of neural networks in politology

dr hab. Igor Podolak (grant coordinated by dr hab. Łukasz Wordliczek)
2015-2018, NCN OPUS 8

Knowledge enriched sparse word embedding

lic. Stanisław Jastrzębski (under supervision of dr hab. Jacek Tabor)
2015-2018, MNiSW Diamentowy Grant 4

Minimal Memory Clustering Paradigm

dr hab. Jacek Tabor, dr hab. Igor Podolak, prof. dr hab. Andrzej Bojarski, mgr Przemysław Spurek, mgr Marek Śmieja, mgr Wojciech Czarnecki, mgr Sabina Smusz
2015-2017, NCN OPUS 7

Development of machine learning methods with applications to chemical compound activity prediction

mgr Marek Śmieja
2015-2017, NCN PRELUDIUM 7

A novel approach to de novo genome assembly problem based on UCT

mgr Ewa Matczyńska
2014-2017, NCN PRELUDIUM 6

The memory center

mgr Przemysław Spurek
2014-2017, NCN PRELUDIUM 5

Novel active learning querying strategy for the machine learning models

mgr Wojciech M. Czarnecki
2014-2017, NCN PRELUDIUM 5

Prizes

Rector prize for scientific achievements

prof. Jacek Tabor, dr Wojciech Czarnecki, dr Przemysław Spurek
2015, UJ

Scholarship of Polish Minister Of Science

lic. Stanisław Jastrzębski
2015, MNiSW

Best paper award at CORES'15

mgr Wojciech Czarnecki and dr hab. Jacek Tabor
2015

4th place in TRADESHIFT competition

lic. Stanisław Jastrzębski, Rafał Józefowicz (Google), mgr Wojciech Czarnecki
2014

5th place in CONNECTOMICS competition

mgr Wojciech Czarnecki, Rafał Józefowicz (Google)
2014

Second prize at the SMP competition

Stanisław Jastrzębski (under supervision of dr hab. Igor Podolak)
2013

Best presentation at CISIM'13

mgr Przemysław Spurek
2013

Projects

GMUM.r

GMUM.r is an R package done in a close collaboration between team members and students. This is an unique occasion to make a first step into true science. We encourage all students to visit a section for students as well as contact Wojciech Czarnecki (project coordinator) in order to get more details.

Github

CRAN

HCOC

Hierarchical Classifier with Overlaping Clusters is a novel ensemble model designed to work with problems with numerous output labels. It is being developed mainly by dr hab. Igor Podolak, polish speaking people may be interested in the new book describing this classifier in detail.

contact

CEC

Cross Entropy Clustering is a novel clustering method developed by dr hab. Jacek Tabor and his PhD students. It finds a compromise between easy and efficient implementation (similar to the k-means) and strengths of the model based clustering (Expectation Maximization). We are currently working on the generalizations of this concept and applications in the classification.

contact

Contact Us

Igor Podolak, PhD
Group of Machine Learning Research
Faculty of Mathematics and Computer Science
ul. Lojasiewicza 6
30-342, Cracow, Poland