Skip to main content
conference lunch move company map contacts lindholmen lindholmen 2 travel info


AI for Solving Real Large-Scale Decision Problems

At AI Innovation of Sweden seminar April 26, we were visited by Emma Frejinger, Université de Montréal, who let us learn more about AI for Solving Real Large-Scale Decision Problems, and the topic "The Powerful Combination of Machine Learning and Discrete Optimization". Here you have the chance to take part of the seminar, watch, learn and enjoy!

The field of Artificial Intelligence (AI) is flourishing and AI technologies are becoming part of people's everyday life. The success can to a large extent be attributed to breakthroughs in machine learning (ML) and in particular deep learning. ML algorithms have had remarkable success in automating tasks that are easy to accomplish but difficult to formalize by humans, for example, image analysis, natural language processing, voice and face recognition.

On the other hand, the field of Operations Research (OR) has been successful in developing methodologies to solve efficiently various types of decision problems that can be formalized but are nevertheless too complex or time consuming for humans to process. These methodologies are crucial to a wealth of applications. 

In this talk Emma gives an introduction to ML and OR with a special focus on novel AI methodologies at the intersection between ML and discrete optimization. She gives a high-level presentation of such a methodology where OR and ML complement each other to address an overall decision problem that could not be solved otherwise. It is applied to a container loading problem faced by North American railroads. 

The talk is based on joint work with Yoshua Bengio, Sébastien Lachapelle, Simon Lacoste-Julien, Eric Larsen and Andrea Lodi.´

About Emma Frejinger
Emma Frejinger holds a Ph.D. in Mathematics and is Associate Professor in the Department of Computer Science and Operations Research at Université de Montréal. Her research activities lie at the 

emma12865_0.jpgintersection between operations research and machine learning with a particular focus on transport applications. She directs a research team of some fifteen students, postdoctoral fellows and research associates on these top ics. Along with her students, she has won several international awards for her research on predicting user behavior in transport networks.

Emma Frejinger is the holder of the Canadian National Railway Company (CN) Chair in Optimization of Railway Operations, she is a member of CIRRELT and an associate member of Mila. She also holds a part-time position as scientific advisor of IVADO Labs, a provider of AI-driven supply chain solutions.