Velibor Misic

What products should a firm launch? What mix of products should a retailer offer? How should a firm leverage data about its customers to make personalized decisions? These are just some of the questions that Assistant Professor Velibor Mišić is interested in.

“Firms nowadays need to make decisions of enormous complexity,” says Mišić. “They simultaneously have more and more data available to guide them in their decisions. My research is focused on developing analytics methodologies that allow firms to transform this data into decisions that create value.”

As an example, Mišić points to product line decisions. “By varying combinations of product features, there could be thousands of possible products, and firms have to select a set of these products to launch, leading to an even larger number of product lines,” he says. “At the same time, the data that tells firms how customers value these attributes is limited and it leads to multiple models of how customers choose. These models are often inconsistent with each other; they lead to different predictions and ultimately imply different product line decisions. One of my research streams has considered how to make good product line decisions in the face of uncertainty about how customers will behave.”

Mišić’s research has spanned a multitude of subjects in the area of analytics, such as choice and assortment problems, robust optimization, dynamic decision making under uncertainty and health care. His research has been published in journals such as Operations Research, European Journal of Operational Research and Computers & Operations Research. He earned his Ph.D. degree at MIT, and master’s and undergraduate degrees from the University of Toronto.

Outside of research, Mišić has also been involved in teaching. At MIT, he was involved in developing the online version of a popular MBA elective known as The Analytics Edge, which has seen a cumulative enrollment to date of over 100,000 students with diverse educational backgrounds from all over the world. He has served as a teaching assistant in the residential version of this class.

“I greatly enjoy teaching,” says Mišić. “I believe teaching can have great impact because students will decide whether or not analytics gain traction in practice. A student who asks me a question about clustering or regression today is someone who might use that knowledge within a company tomorrow. The recognition of the potential for this kind of impact is a major force that guides how I interact with students.”

Education

• Ph.D. Operations Research, 2016, Massachusetts Institute of Technology
• M.A.Sc. Industrial Engineering, 2012, University of Toronto
• B.A.Sc. Industrial Engineering, 2010, University of Toronto

Working Papers

Exact Logit-Based Product Design. Akçakuş, İ., and Mišić, V. V. (2021) Submitted.

Assortment Optimization under the Decision Forest Model. Chen, Y.-C., and Mišić, V. V. (2021) Submitted.

Column-Randomized Linear Programs: Performance Guarantees and Applications. Chen, Y.-C., and Mišić, V. V. (2020) Major revision in Operations Research.

Refereed Journal Articles

Decision Forest: A Nonparametric Approach to Modeling Irrational Choice. Chen, Y.-C., and Mišić, V. V. (2021) Forthcoming in Management Science.

A Simulation-Based Evaluation of Machine Learning Models for Clinical Decision Support: Application and Analysis Using Hospital Readmission. Mišić, V. V., Rajaram, K. and Gabel, E. (2021) npj (Nature) Digital Medicine, 4 (98): 1-11.

Interpretable optimal stopping. Ciocan, D. F., and Mišić, V. V. (2019) Forthcoming in Management Science.

Machine Learning Prediction of Post-Operative Emergency Department Hospital Readmission. Mišić, V. V., Gabel, E., Hofer, I., Rajaram, R., and Mahajan, A. (2019) Anesthesiology, 132 (5): 968-980.

Optimization of Tree Ensembles. Mišić, V. V. (2020) Operations Research, 68 (5): 1605-1624.

Data analytics in operations management: a review. Mišić, V. V. and Perakis, G. (2020) Manufacturing & Services Operations Management, 22 (1): 158-169.

Exact first-choice product line optimization. Bertsimas, D., and Mišić, V. V. (2019) Operations Research, 67 (3): 651-670.

The airlift planning problem. Bertsimas, D., Chang, A. A., Mišić, V. V., and Mundru, N. (2019) Transportation Science, 53 (3): 773-795.

A comparison of Monte Carlo tree search and rolling horizon optimization for large scale dynamic resource allocation problems. Bertsimas, D., Griffith, J. D., Gupta, V., Kochenderfer, M. and Mišić, V. V. (2017) European Journal of Operational Research, 263 (2): 664-678.

Robust product line design. Bertsimas, D., and Mišić, V. V. (2016). Forthcoming in Operations Research.

Decomposable Markov decision processes: a fluid optimization approach. Bertsimas, D., and Mišić, V. V. (2016). Forthcoming in Operations Research.