The following paper was written February 4-8, 2021 with the help of two other teammates as we participated in The Interdisciplinary Contest in Modeling. We chose to answer Problem D: The Influence of Music. Using data scraped from Spotify that contained songs from the 1930's to the present day, we were tasked with determining whether or not certain artists influenced others, what defined genres of music and how genres changed over time.
We implemented a variety of tools to solve this problem, most notably using a neural network and metrics supplied in the song data, like tempo or loudness, to predict whether or not a given song was of a given genre, which was fairly successful, and using a neural network to predict whether or not a given artist was influenced by another artist, which was not as successful.
We did all of our programming in R and you can find the data that we used and the code and graphs we created here.
Our submission was awarded the designation of "Honorable Mention". This placed us in the top 30% of submissions for the problem we chose. Official results can be found here.