An interview with Georgia Institute Technology’s Associate Professor, Dr. He Wang
In this case study, Dr. Wang discusses his successful integration of Vocareum Notebook into his AI and Data Analytics course. Dr. Wang emphasized Vocareum’s user-friendly approach, particularly its efficient GPU access for machine learning training, simplifying complex setups for students. He praised the positive student experience, highlighting the platform’s utility in preparing them for real-world applications. With plans for course expansion and potential online offerings, Dr. Wang commended Vocareum’s scalability and accessibility, recommending it to colleagues for its effectiveness in enhancing the learning environment.
Dr. He Wang, Associate Professor at Georgia Institute of Technology in the Industrial and Systems Engineering Department
Dr. He Wang is the Jerry and Harriet Thuesen Early Career Professor and an Associate Professor in the H. Milton Stewart School of Industrial and Systems Engineering.
His research focuses on the interface between machine learning and operations management, where he develops data-driven methods for applications including supply chain management and dynamic pricing.
He received his Ph.D. in operations research and his M.S. in transportation from Massachusetts Institute of Technology. He received his B.S. in industrial engineering and math from Tsinghua University in China.
I’m an Associate Professor at Georgia Tech in the Industrial and Systems Engineering department. I’ve been teaching here since 2017 and hold a PhD in Operations Research.
I was introduced to Vocareum Notebook by a colleague from the School of Computer Science. I was preparing a new course on data analytics and needed a platform that allowed students to work with Jupyter notebooks effectively.
The course aims to bridge the gap between basic Python programming courses and more advanced topics like machine learning. I noticed that students often lack experience in handling real-world, unclean data, so the first half of the course focuses on teaching Pandas for data manipulation and libraries for data visualization, followed by machine learning tools in the second half.
Vocareum’s integration with our LMS (Canvas) has been seamless. Everything from lectures to assignments is managed through Jupyter notebooks. For instance, I can post a link in Canvas that directly opens the notebook in the student’s browser. This uniform environment is incredibly efficient. We also utilize nbgrader (a Jupyter Notebook extension) for auto-grading assignments in Vocareum, allowing students to receive real-time feedback for coding questions. Another nice feature of Vocareum is that grades are automatically synced with the LMS.
GPUs have become essential for machine learning training. We use them to demonstrate their speed advantage over CPUs for handling large training sets. Students run models on both CPU and GPU to compare performance. For this, we rely on Vocareum for its straightforward access to GPU resources.
While we have GPUs on campus, accessing them is not always user-friendly. Vocareum makes the entire process smoother. Students don’t need to worry about complex setups or command lines; they can just open a notebook, and everything is ready.
The student experience has been quite smooth. They appreciate not having to deal with the setup complexities. As an instructor, I can manage package installations for students. In the industry, companies often manage these setups when they provide cloud-based computing environments to their employees, so it’s great that Vocareum provides a similar environment that allows students to experience what they may use for their future jobs.
I think I’ve benefited a lot from the product. I’ve recommended this to many of my colleagues and friends. I plan to expand the capacity of this course and potentially offer it online. The scalability and accessibility of Vocareum make it an ideal platform for this expansion.