Staff: N. Antulov-Fantulin, I. Ivek, M. Mihelčić, R. Horvat, D. Gamberger, M. Piskorec,  I. Maric, T. Smuc

Laboratory for Machine Learning and Knowledge Representation

Development of machine learning algorithms and their applications in data mining and knowledge discovery tasks. 

Tomislav Šmuc


PhD Tomislav Šmuc

+385 1 456 1085
Fax: +38514680114

Moira Španović

first wing, room 103

phone and fax  +385 1 4680114

left right

 The Laboratory works on scientific research in the field of technical sciences and specifically on:

  • research and development of machine learning algorithms and their application in data analysis
  • theory and application of knowledge representation and reasoning algorithms
  • development of numerical algorithms for modeling of complex systems

Currently we work on:

  • development of algorithms for rule induction from large datasets
  • evaluation of outlier detection algorithms and development of a new approach based on random forests
  • data mining in social applications (political instability, direct foreign investment, international tourism industry)
  • detection of subgroups of patients by mining gene expression data
  • medical and technical ontologies as well as representation and reasoning for procedural knowledge
  • evaluation of the quality of computational annotations in the gene ontology
  • development and application of the GMDH based modeling
  • development of surrogate models of the reduced complexity for embedded computer systems

Ivan Marić
dr. sc.

Senior scientist
+385 1 456 1191
Konstrukcija GMDH modela

Computational Intelligence Methods in Measurement Systems

Glavni istraživač / voditelj: Ivan Marić

The goal of the project is to investigate the possibilities of development and application of the artificial intelligence methods and machine learning techniques in measurement systems ranging from the development of simple models to the synthesis of sophisticated measurement procedures and smart instruments capable not only to execute simple commands but to perform rather complex tasks. The research will be focused on algorithms and methods for reducing the processing time and the complexity of measurement procedures and for increasing the accuracy, flexibility and the reliability of both embedded and distributed measurement systems.

more »

This is our main machine for complex computations. 

At the moment the most powerful server in the Laboratory is Crunchy. His main characteristic is 4 processors with 8 cores each working on 2.4GHz, meaning that at the same we can execute 32 processes. Working memory is 32 GB. Disk space is 500 GB. Operating system is Linux Debian. We are proud on this machine because it has been completely assembled in our lab. The credit belongs to Rajko Horvat.

In the Lab data storage is served by machine Qnap with 8 TB disk space in RAID 1 mode.



book Rule Learning

Book: Foundations of Rule Learnig
authors: J.Fuernkranz, D.Gamberger, and N.Lavrac

Rules – the clearest, most explored and best understood form of knowledge representation – are particularly important for data mining, as they offer the best tradeoff between human and machine understandability. This book presents the fundamentals of rule learning as investigated in classical machine learning and modern data mining. It introduces a feature-based view, as a unifying framework for propositional and relational rule learning, thus bridging the gap between attribute-value learning and inductive logic programming, and providing complete coverage of most important elements of rule learning.   The book can be used as a textbook for teaching machine learning, as well as a comprehensive reference to research in the field of inductive rule learning. As such, it targets students, researchers and developers of rule learning algorithms, presenting the fundamental rule learning concepts in sufficient breadth and depth to enable the reader to understand, develop and apply rule learning techniques to real-world data.