Hanbo Yu

Lecturer, Administrative Sciences

Lecturer Hanbo Yu brings more than a decade of experience to the classroom, spanning data visualization, statistical modeling, machine learning, and deep learning, as well as production workflows in R, Python, SQL, and MATLAB. He previously worked as an energy market analyst/asset manager for Agilitas Energy (Boston area) and as a data scientist in the investment department of the State Grid Corporation of China (Beijing), where he developed forecasting, trading, and risk analytics for large-scale energy and investment operations. A graduate of Boston University’s Master of Science program in Applied Business Analytics, Yu was recipient of BU MET’s Award for Excellence in Graduate Studies in 2021. He has contributed to the curriculum development across MET’s analytics offerings, including course development support for Python and SQL for Business Analytics (MET AD 599) and related pre-analytics laboratories.

His current scholarly work develops Signal Interpretation Economics (SIE), a computational research program that models economic outcomes as a pipeline of signals, interpretation, belief anchors, feasibility constraints, and behavioral responses, with applications in macro-finance (interest rates, asset pricing), crises (bubbles/hyperinflation, depressions/stagflation), political economy, and reliable AI-agent design.

Research Interests

Courses

Faculty Q&A

ABA Faculty and Alum’s Research Delves into Deep Learning with Neural Networks: “The Middle Area Between Machine Learning and AI.”

What is your area of expertise?
I specialize in finance, financial trading, energy markets in the US, machine learning, probability theory and statistics, and neural networks.

Please tell us about your work. Can you share any current research or recent publications?
My recent research involved financial trading in the energy market, and a paper was published in the Proceedings of the Conference on Software Engineering and Computer Science 2023 (CSECS 2023).

What are you currently working on?
I am working on multiple research projects, including one concerning deep learning with neural networks, a business simulation research project, and an analytics lab research project.

How does the subject you work in apply in practice? What is its application?
Deep learning with neural networks is the middle area between machine learning and AI—its applications include financial trading, time series analysis in the stock and futures markets, image recognition, and, of course, regression/classification analytics. It can be widely applied in a lot of areas.

What courses are you or will you be teaching at MET? What program are you teaching in? How does your past experiences help to inform your teaching?
I coordinate and teach Introduction to Python and SQL for Business Analytics (MET AD 599) and Enterprise Risk Analytics (MET AD 616). I also helped teach the Capstone Project in Applied Business Analytics (MET AD 899), and supervise two groups of students with two different research projects. I am teaching in the Applied Business Analytics (ABA) program. I graduated from the ABA program as well, and before the ABA program, I had about 10 years of teaching experience. Therefore, I know the courses well enough, and I am trying to find space for improvement for the courses. My experience helped me a lot in my current job.

Please highlight a particular project within one of your courses that most interests your students. What real-life exercises do you bring to class?
Before I joined the MET faculty full time, I was a quantitative trader focusing on the energy market in the US. I shared my trading experiences (quantitative analysis such as deep learning/machine learning models, qualitative analysis such as market participants behavior/overreaction/correction, etc.) with one group of students in the Capstone Project, and gave them a target rate of return of 30 percent within 3 months. In my previous work, for the financial trading part, I reached over 400 percent rate of return annually, therefore, 30 percent was not too unrealistic for the students. They were excited about this opportunity, and they reached the target within 45 days. That was a real-life exercise I brought into class, and it worked quite well.

View all profiles