Sule Balkan & Tim Anderson
Energy Demand Forecasting
This class was a special reduced class time hands on course that centered around the Global Energy Forecasting Competition which was an international forecasting competition hosted Tao Hong using real data on electricity demand. The New England region was the point of origin for the real data and was broken up into several regions including: Maine, New Hampshire, Vermont, Connecticut, Rhode Island, 3 regions of Massachusetts, and the combined region ISO NE CA. The competition was broken up into two brackets based on the amount of data that the participant intended to use which was either two years of data or an unconstrained amount of data so that the participants could utilize all presently available data. The participants could also choose to work in a team or not. I choose to work with my fellow classmate, Mohamed Almusallam, with whom I had collaborated successfully in the past. We choose to utilize the constrained data set of two years.[This report mentions this collaboration solely so that the reviewer of this report will take cognizance that all efforts in this report share only a partial account of the work done this term and represent in most cases my own first attempts to create energy forecasts that in the later cases( especially the neural network) were further improved with help of my collaborator]. With that mentioned, this report will summarize and examine the energy forecast attempts that I put forth over the term.
Davis, Michael, "Energy Forecasting" (2017). Engineering and Technology Management Student Projects. 103.