Objective

​

Our task is to predict the winner of the 2017 GRAMMY Record of the Year award. Because our model can be easily applied to future GRAMMY seasons, these results will yield insights into subtle differences between popular expectations and actual winners, especially in upset years, and can help provide context for identifying shifts and trends in popular music. Our project output is a ranking of all songs that have made an appearance on the Billboard Top 100 in this year’s GRAMMY eligibility period by their likelihood of winning the Record of the Year award.