Penn State Supercomputing tackles jazz improvisation

Love it or hate it, improvisation – while it may seem random – is often more focused and structured than it seems. And, as unlikely as it may seem, supercomputing is also at play here: A team of researchers led by Penn State applied the university’s Roar supercomputer to map and analyze patterns in the jazz greats’ improvisational music. all time. The goal: to understand the methods and motives that underlie creativity.

The research began with the Weimar Jazz Database, which contains compositions and improvisations by renowned musicians like John Coltrane and Miles Davis. The research team approached this data from a syntactical and linguistic perspective, breaking the improvisations down into more manageable chunks.

“We’ve largely thought of improvisation as something called hierarchical behavior,” Hannah Merseal, a Penn State graduate student and member of the Cognitive Neuroscience of Creativity Lab, explained in an interview with Penn State’s Matt Swayne. something that happens in real time and is organized such that smaller subunits report to larger units, which is also kind of how language works. We have a sentence, then smaller pieces of the sentence below.

These tracks, she explained, are central to the improvisation process.

“When we think of improvisation, a lot of us think of players picking up smaller pieces of music – like musical notes that they have stored in memory,” she said. “They aggregate this information in ways that are easy to store and reorganize.”

The researchers noticed that the sequences found in the improvisations in the Weimar Jazz Database increased in complexity with each phrase, illustrating an “easy-first” principle. They analyzed the solos in sequences of five notes, then recruited participants to quantify the relationship between these sequences.

Modeling the resulting data was a computationally strenuous task – but luckily the team had access to Penn State’s Roar supercomputer, which contains over 1,000 servers and over 23,000 cores, as well as 18 PB of storage. The power of Roar has allowed researchers to model and analyze the network, for example by measuring semantic distance – a common metric for creativity that relies on the conceptual distance between words and sentences.

“Our memories can organize categories of things by grouping them together and we can model that as a network,” explained Roger Beaty, assistant professor of psychology at Penn State and principal investigator of the Cognitive Neuroscience of Creativity Lab. “Previously, there was research on network modeling and semantic memory. More recently, we and some of our colleagues have tried to test this using computational approaches to network science.

The researchers found that jazz greats tended to start with simple, familiar sequences, then slowly evolve into more complex creative choices later in the solos.

To learn more about this research, read the report by Matt Swayne of Penn State.

About Mariel Baker

Check Also

AMD Releases Latest Consistent Device Memory Mapping Linux Code – Designed for Frontier

Over the past year, we have seen various patches released by AMD engineers with a …