At the Machine Learning Center at Georgia Tech (ML@GT), we’re lucky to have over 150 faculty members affiliated with our mission. With faculty from all six colleges on Georgia Tech’s campus plus several from Georgia Tech Research Institute (GTRI) and the Enterprise Innovation Institute (EI2), our professors bring a broad range of skills and areas of expertise.
Pascal Van Hentenryck is no exception. Van Hentenryck is the A. Russell Chandler III Chair and Professor in the H. Milton Stewart School of Industrial and Systems Engineering at Georgia Tech. His research focuses on artificial intelligence, data science, and operations research. His current focus is to develop methodologies, algorithms, and systems for addressing challenging problems in mobility, energy systems, resilience, and privacy.
In the past, his research focused on optimization and the design and implementation of innovative optimization systems, including the CHIP programming system (a Cosytec product), the foundation of all modern constraint programming systems and the optimization programming language OPL (now an IBM Product).
Widely recognized and awarded for his accomplishments, Van Hentenryck is a fellow of the Association for the Advancement of Artificial Intelligence (AAAI) and the Institute for Operations Research and Management Science (INFORMS.) His work received the Test of Time Award from the Association of Logic Programming and numerous best paper awards, including from IJCAI and AAAI.
Tell us about your hobbies and education.
I come from Belgium, a small country in Europe. I went to college to study economics but I took a class in computer science and I immediately became addicted to it and I was fortunate to have some incredible professors. I spent a lot of time in the lab, coding on a DEC-20 machine initially, before we could get access to PCs and Macs. Sometimes at night, I had to jump out of the windows to get out because the all doors of the computer science institute were closed.
Outside work, I run a lot and I love to travel, read, and watch movies (Netflix series these days).
What do you research and what about it interests you?
I spent the first part of my career building programming languages and optimization systems (I probably wrote well over 1,000,000 lines of code in my life, mostly in C++). With these systems, I could write compact programs or models that could solve really complex scheduling, routing, and resource allocation problems.
Nowadays I use optimization and machine learning to solve complex societal challenges, primarily in energy, mobility and transportation, and privacy.
In energy, the challenges are to move to a grid with a substantial share of renewable energy through the combination of machine learning and optimization. In mobility, we are looking at a new generation of mobility systems powered by data science, machine learning, and optimization. What is unique about these research areas is the combination of challenging scientific problems and the potential impact on society.
What drew you to your field of study and to being a professor?
After my Ph.D., I had to choose between joining a startup, going to industry, or continuing to do research. One of my American friends told me to apply to the US. He gave me the list of the top 20 schools in computer science and told me to apply there. Then I got lucky.
What has been the biggest challenge about being a professor?
Thinking does not have a 9-5 schedule…so it is hard to balance work and family time. When our kids were younger and we were having dinner together, I would sometimes start to think about a bug, an idea, or a better way to explain something. Our children would recognize this instantly and would just stop talking until I came back to earth. I am really bad at this balancing act and I am incredibly lucky that my wife has a computer science degree.
Why did you choose Georgia Tech?
There is something unique about Tech. There are so many faculty who are at the top of their fields and all this expertise is there for you to experience in so many disciplines. It is really humbling. The students are incredibly dedicated to their craft. They are really fun to teach and work with.
What do you enjoy about machine learning and being affiliated with ML@GT?
Georgia Tech has an amazingly strong and diverse group of faculty and students, bridging so many areas of ML, from deep theoretical results to applications across many disciplines. ML is old but it really has been rejuvenated and there is so much left to explore. You get this sense of excitement and a taste of all these opportunities here at Tech through ML@GT.
You run a computer science summer camp every year at Georgia Tech. Tell us about it!
I run the Seth Bonder Summer Camp on Computational and Data Science. It was held at the University of Michigan in 2017 and 2018 before coming to Georgia Tech in 2019. Due to the pandemic, it will be held online this summer.
It’s intended for high school students with little to no experience with computer science and programming. It covers basic computing concepts and applies them to various disciplines like social science, machine learning, and computational biology.
I have always believed that every student in high school should be exposed to computing in a way that makes it interesting even if they are interested in social sciences and the humanities. The camp was born out of testing this idea. I designed and taught a course on this while running research for the Australian Research Center, and when I returned to the United States, I decided to turn it into a summer camp for high school students.
I love running the camp because I just love computer science, machine learning, and optimization. If the camp changes the life of a few of them by exposing them to this material, it is time well-spent. I am really keen to see how the online version will be received.
What is some advice you would give to someone looking to go into academia as a career?
I am a big believer in letting students experience research and teaching, so that they can judge for themselves. I let them join my research group early in their careers and encourage them to become teaching assistants. I also tell them that no two professors are the same: There are many ways to do research and teach.
What is your proudest accomplishment so far?
I have built a few optimization systems that have been in use in academia and industry for more than 20 years. It is pretty cool when you can move from a theoretical idea and transform it into a tool that is widely used across the world. It is also very scary. I also love to teach and I am proud when people tell me that they like my Coursera class.
Tell us something about you that would surprise us.
I played soccer in front of a larger crowd than the audience in any of my keynotes (including IJCAI and NeurIPS). That tells you something about my soccer and academic skills 😊
What is something you wish you knew at the beginning of your career that you know now?
One of my friends (Michela Milano) told me once that I was completely wrong on grant writing. She told me that this is an opportunity to learn new things, to discover new areas, and to think. She was much younger than me but she had it all figured out.
Why do you think embodying Georgia Tech’s motto of “progress and service” is important, especially in regards to artificial intelligence and machine learning?
AI and ML are like space exploration. It is an exciting field with many possibilities. It inspires students, captures the imagination of the public, and raises fascinating scientific and technological challenges. But ultimately we also need to think about how it can benefit society at large and address the fundamental societal challenges we face.
Academics are in a unique position to work on technically interesting problems with substantial social impacts and AI/ML now presents an incredible opportunity to do so in many disciplines. So scientific progress in service of society!
mlatgt.blog 