The purpose of this project is to create a useful tool for musicians that utilizes the harmonic content of their playing to recommend new, relevant chords to play. This is done by training various Long Short-Term Memory (LSTM) Recurrent Neural Networks (RNNs) on the lead sheets of 100 different jazz standards. A total of 200 unique datasets were produced and tested, resulting in the prediction of nearly 51 million chords. A note-prediction accuracy of 82.1% and a chord-prediction accuracy of 34.5% were achieved across all datasets. Methods of data representation that were rooted in valid music theory frameworks were found to increase the efficacy of harmonic prediction by up to 6%. Optimal LSTM input sizes were also determined for each method of data representation.

In the early history of jazz, many of the songs that were popularized by jazz musicians became known as jazz standards, and these songs remain a central component of the jazz repertoire today. Many of these jazz standards were adapted from early Broadway musicals and revues. Rodgers and Hammerstein's Oklahoma! is widely considered to be the most significant musical in the history of Broadway theater. Its innovative blending of song, dance, and a cohesive dramatic story has profoundly influenced the structure of Broadway musicals to this day. However, none of the songs from this show have risen to the status of a jazz standard, and many appear to have not been adapted to the jazz idiom at all. In my Barrett Honors creative project, I have reimagined and arranged nine songs from the original Rodgers and Hammerstein production. I then led a jazz quintet through two months of rehearsals, culminating in a performance of the work for my senior jazz performance recital. A link to the performance at the ASU School of Music Recital Hall is included here: https://www.youtube.com/watch?v=jeOs4muj12M