Zoran Dzunic

(zoki AT mit DOT edu)


* Please find my CV here.

I am a PhD candidate in John Fisherís SLI (Sensing, Learning, and Inference) group at CSAIL at MIT, working on Bayesian structure inference and interaction analysis.

In particular,
I consider the problem of Bayesian inference under uncertainty of evolving temporal graphical structure describing the interaction among multiple time-series. Possible application domains include video analysis, social networks, games, finance, biology, neuroscience, climate, sensor data, etc.

My general research interests lie in machine learning and statistical inference, particularly graphical models, sampling algorithms, and time-series analysis. In the past, I have also worked on natural language processing (NLP) problems.

I received my master's degree from
MIT in 2009, and my bachelor's degree in computer science from the University of Nis in 2005. I interned at Microsoft Research in 2009 and at Mitsubishi Electric Research Laboratory (MERL) in 2010. Prior to coming to MIT, I worked as a software engineer in a Data Mining Team in Accordia.

I TA-ed Machine Learning (6.867) in Fall '08 and Introduction to Algorithms (6.006) in Fall '09.

In the past, I participated in programming contests and am passionate about algorithms and algorithmic problems. I won silver medal in the International Olympiad in Informatics (IOI) in 1999. I later coached students in Serbia, and authored a problem for IOI 2001. I was also one of the coaches of MIT ACM-ICPC team in 2007-08 and 2008-09.

Recent publications:

Z. Dzunic, J. G. Chen, H. Mobahi, O. Buyukozturk, and J. Fisher III. A Bayesian State-space Approach for Damage Detection and Classi cation. In Proceedings of the Twenty-Third International Modal Analysis Conference, to appear.

Z. Dzunic and J. Fisher III. Bayesian Switching Interaction Analysis Under Uncertainty. In Proceedings of the Seventeenth International Conference on Arti cial Intelligence and Statistics, pp. 220-228, 2014. [bib]