Students in the course Energy Analytics have been competing very successfully in international energy forecasting competitions for a number of years. This fall, the course instructor decided the Lee College of Engineering would host a competition of its own, and 142 data scientists from 26 countries have formed 81 teams for the first UNC Charlotte BigDEAL Forecasting Competition 2018.
Course instructor Tao Hong, associate professor and director of EPIC Big Data Energy Analytics Laboratory (BigDEAL), said, “I was so impressed by the Energy Analytics class of 2018 that rather than giving them the same assignments from previous years, I wanted to bring them a bigger challenge. So we have created our own competition and opened it to competitors from around the globe. My hope is my students can test their analytic skills against data scientists around the world and learn something along the way.”
This is the fifth time UNC Charlotte has offered Energy Analytics, a hybrid course mixing both online and on-campus students who are in the Master of Science in Engineering Management program. The course has received the Student Forecasting Award from the International Institute of Forecasters, recognizing the top students from each class offering. In previous years, students from the course have won many awards and top places at international forecasting competitions, such as Global Energy Forecasting Competition 2014 and RWE npower Forecasting Challenge.
The in-class competition offers no monetary prize. During the past two weeks, 142 data scientists from 26 countries formed 81 teams to join the new BigDEAL Forecasting Competition 2018. The qualifying match started Oct. 22, with teams being given three years of hourly load and temperature history and just the temperature history for a fourth year. The contestants are asked to forecast the hourly load of the fourth year.
“In this big data era, we teach our students the skills to extract actionable insights from data,” said Simon Hsiang, chair of the Department of Systems Engineering and Engineering Management. “We have been placing a strong emphasis on data-driven decision making in our curriculum. Forecasting is the best way to connect the dots of our curriculum, to mitigate the risks, to weigh our options and to understand our desires and fears.”
A leaderboard of the qualifying match will be published on Nov. 5. The teams outperforming the bottom-ranked in-class students will advance to the final match; the topic has not been released. The winners of the final match will be announced on Dec. 3.
Image: A world map that indicates the 26 countries competing in the BIGDeal forecasting competition.