Prof. Maxime Cohen Launches New Guide to Demand Prediction in Retail

Book co-authored with two McGill MMA alumni, Arthur Pentecoste and Paul-Emile Gras as well as Dr. Renyu Zhang of NYU Shanghai

In collaboration with Dr. Renyu Zhang, Professor of Operations Management at NYU Shanghai, and Arthur Pentecoste and Paul-Emile Gras, two graduates of the McGill Desautels Master of Management in Analytics (MMA), Prof. Maxime Cohen’s new book, Demand Prediction in Retail – A Practical Guide to Leverage Data and Predictive Analytics is scheduled to hit shelves on January 3, 2022.

We wanted to build a case study based on real-world experience for a lecture, to teach students how to predict demand,” Cohen explained. “There was so much material, the case study was going very well and was 60, 70, 80 pages long. I told [Arthur and Paul-Emile] ‘Look, unfortunately, this is not a case study anymore. I think that this is becoming a book.”

Working with Prof. Cohen as research assistants when the project began, Pentecoste and Gras were able to work closely with the professor till to the end, even once they had graduated and left their roles as research assistants, moving onto careers as data scientists.

“The way it would work typically is we would collaborate with [Prof. Cohen] on how we were going to approach a problem, and the different steps, and we would just receive guidance and help troubleshoot,” said Pentecoste. “And in the meantime, there would be some heads down work on our end where we were just implementing several algorithms and working on the pure coding.”

While there was one target for the teaching case study that was originally planned, the new book format meant that the target audience could be expanded upon. Cohen explained that the book is not only aimed at students in programs like McGill’s MMA, but also at data scientists and retailers who might hire scientists to implement the tools and processes outlined in the book.

The goal of the book is ultimately to demystify how predictive analytics can be used in a retail context and how retailers can monetize historical data, to do better in the future.

“The book includes detailed implementations, along with comprehensive Jupyter notebooks. [There is also] a data set available for download with the book. In fact, the book is accompanied by the website” Cohen said.

What was perhaps most helpful for Cohen was the fact that Pentecoste and Gras fell right into that target demographic, allowing them to provide valuable insight on the digestibility of the methods outlined in the book.

“[We were able to] explain some algorithms with words that were simple and more accessible,” Gras said. “I mean, we both have engineering and math backgrounds in our undergrad, but most of the models covered in the book, two years ago, we did not know. It was [fresh] in our heads, what would have helped us better understand all the topics.”

Both former students appreciated the opportunity to work with Cohen and their other co-author on this project, citing it as a very important source of hands-on experience headed into the workforce.

“I think the fact that it's a professor that has done a good amount of publishing papers and that has been recognized, he has credibility, it helped with my resume and with screening processes,” said Pentecoste. “As a research assistant, because we’ve done some consulting with some companies, we’ve done some research, it was a way to show proficiency in analytics and using those tools.”

It seems the feeling was mutual, as Cohen expressed how much he enjoyed working with Pentecoste and Gras.

My experience working with Arthur and Paul-Emile was great. I’m very happy with my collaboration with them,” Cohen said. “They were very responsible and eager to contribute to the book. They were the right judges to confirm the best way to tailor the content of the book for the target audience. They also had quite a big role in terms of making the code top notch. They took a lot of time to make sure that all the codes and notebooks are well written and easy to follow.”

With the publishing of this new book and his recent nomination as the Scale AI Research Chair in Data Science for Retail, it’s been an exciting time for Cohen, who plans to continue his work on demand prediction and retail analytics, while examining the challenges presented by COVID-19.

“I'm also hoping to branch out and work on several other topics,” Cohen explained. “The Chair will allow me to work on developing data science techniques for retail applications. We also want to make sure that all the methods have social impact—sustainability considerations will be at the core of the Chair. Another focus would be potentially emerging markets [and the] circular economy.”


Back to top