The purpose behind this project was to identify a music artist based on their own music. Since most of our group members are avid listeners of music, we decided that it would be interesting to do a project that would identify a music artist based on their music alone. Not to mention we wanted to solve the annoying occurrence of hearing a good song but not knowing the artist.

​

To achieve this broad goal, we utilized the signal processing techniques along with applied machine learning in order to be able to develop a predictive model to be able to identify the music artist based on their voice. We wanted to achieve this by getting the representation of our data in the frequency domain using the FFT algorithm, filtering out any audio signals at undesirable frequencies, and fitting the processed data into machine learning models to then predict the music artist based on the song. 

​

Due to the short nature of this project, we started out with samples from 10 songs from three different music artists each: Taylor Swift, Drake, and Pitbull. The objective is to be able to process the audio from each of these music artists and be able to create machine learning classifier models to identify the music artists. The goal was for the classifier to be able to accurately identify the music artist in approximately 75% of the testing samples or higher.