BANJA LUKA SCIENTIST PUBLISHED IN LEADING ARTIFICIAL INTELLIGENCE JOURNAL

Dr. Marko Đukanović, Assistant Professor at the Faculty of Natural Sciences and Mathematics in Banja Luka, has co-authored a groundbreaking research paper featured in one of the world’s most prestigious journals in the field of artificial intelligence — Expert Systems With Applications, which boasts an impressive impact factor of 7.5.

The paper, titled "Learning Search Algorithm for the Restricted Longest Common Subsequence Problem", introduces an innovative approach showing how artificial intelligence can assist in solving complex problems in bioinformatics.

The research brings together two powerful tools of modern science: heuristic algorithms, which allow efficient navigation through vast search spaces, and neural networks, computational models inspired by the way the human brain learns. By combining these technologies, the researchers developed a hybrid system capable of making improved decisions within the search space, resulting in significantly higher-quality solutions.

A particularly valuable aspect of the study is its use of explainable artificial intelligence — a method that provides clear reasoning behind the algorithm’s performance. This capability to interpret why an algorithm performs better or worse based on the attributes of a given problem instance is what makes this work especially impactful. In fields like combinatorial optimization, where algorithm decisions are often treated as black boxes, such insights represent a major advancement in improving algorithm efficiency and transparency.

The research was conducted through a collaboration between six scientists from Slovenia, Serbia, Spain, and Bosnia and Herzegovina. Alongside Dr. Đukanović, the co-authors include Jaume Reixach and Dr. Christian Blum from the Artificial Intelligence Research Institute in Barcelona, Ana Nikolić and Dr. Tome Eftimov from the Jožef Stefan Institute in Ljubljana, and Prof. Dr. Aleksandar Kartelj from the Faculty of Mathematics at the University of Belgrade.

The project was supported by a bilateral scientific collaboration between the University of Banja Luka’s Faculty of Natural Sciences and Mathematics and the Jožef Stefan Institute. It also received funding from the Ministry of Scientific and Technological Development and Higher Education of the Republic of Srpska through the national project “Application of Deep Learning to Improve Optimization Algorithm Efficiency”, led by Dr. Đukanović.