Welcome to my personal page!

Image credit: Unsplash


Hi everyone!

I decided to create a personal website to show what is happening in my work and research life. There has been a lot of ups and downs, but I’m proud of what I’ve achieved in previous years.

I’ll try to keep the website as accurate as possible, with the latest info regarding work that I’ve done.

Latest News

The latest info I have is that I have become Assistant Professor at University of Belgrade. I’ll be teaching on subjects Decision theory, Business Intelligence, Decision support systems, and Machine Learning on bachelor level of studies, as well as on subjects Business Intelligence systems, Data Warehouses, Data Mining, Design of Machine Learning algorihms on master level of studies. I forgot to mention subjects Data Science, Advanced Machine Learning and Bioinformatics! (a lot of subjects to teach)

Also, it is a pleasure to inform you that my paper FairDEA - Removing Disparate Impact from Efficiency Scores got accepted in the European Journal of Operational Research (EJOR)! The review process lasted for 16 months (six months are on me), and paper got out ~15 pages “stronger” after the review.

(Very brief) Research summary

Since the aim of this page is to showcase my research, I’ll briefly explain what are my research directions. They are mostly related to research projects I’m involved with. They are sorted in descending order (from the most recent one to the oldest).

  • Fairness and justice in algorithmic decision making This is the area that, currently, interest me the most. The idea is to alleviate (or completely eliminate) biases that exist in the real world when one employs machine learning models. This part of my research refers to the various attempts at correcting algorithmic bias, mostly in classification settings. However, we have some papers that tries to equalize the outcome between the groups in decision making methods, or data envelopment analysis. This project is funded by Office for Naval Research.

  • Ski injury predictions When we were between two projects (one above and one below this one), we spent time trying to understand and explain ski injury phenomena on Mt. Kopaonik. Since some members of our team have a lot of personal experience (as a moutain ski rescue service), we wanted to create data driven models that predicts where and when the injury occurs. This information could help rescue service to better prepare for the (unwanted) event.

  • Machine Learning for Medicine My first sail in research (project related) waters was when our research team got a project from the Swiss National Science Foundation. As a result of this project, we published several papers where we designed algorithms for the task of hospital readmission prediction.

  • Honorable mention White Box Algorithm Design As a student, I started working on the WhiBo - white box algorithm design. More specifically, each algorithm can be thought of as a composition of subproblems that needs to be solved. The idea was to “dissect” algorithms into independent components, analyze the effects, and use them as needed. We created a RapidMiner extension (for decision trees) and R package (for clustering). I gave a talk at SatRday regarding Whibo clustering!