Grant Gollier
The dataset contains product demand over a span of several years and came from here on Kaggle. I will try to include it in the repository, but if Github doesn't let me upload it just download the dataset and rename the csv to Historical_Product_Demand.csv and place it in the data directory.
Overall, I think my results were pretty good. However, both models I tried didn't do a great job of handling outliers which I guess is to be expected since those are most likely due to demand spikes that wouldn't be due to date alone, but maybe some other outside factor such as a sale or ad campaign. To see a more indepth discussion please look at the Time Series Notebook.