PLAINVIEW – Taking part in Texas Undergraduate Research Day at the state capitol, Wayland Baptist University senior Ilan Joffe presented his work today on using clustering algorithms to identify similar music pieces.
Using a research poster, Joffe provided a brief overview of his undergraduate research project, “Does Genre Mean Anything: Classifying Music with Artificial Intelligence.” He presented his findings and answered questions about his research.
Joffe’s poster was among more than 55 research projects from 46 private and public universities featured during Texas Undergraduate Research Day.
“Music is notoriously difficult to classify,”Joffe said in his research project abstract. “It is an integral part of the human experience, and yet each person has their own impressions, preferences, and opinions on what makes music great. Typically, music is classified using genres. In this work, we explore an alternative classification method.”
Joffe’s research uses clustering algorithms to identify similar music pieces based on waveforms as well as features extracted from waveforms.
“We compare the performance of a variety of artificial intelligence algorithms in classifying music into both new and pre-existing groups,”Joffe stated in the abstract. “Using Regression, Decision Trees, and Artificial Neural Nets (ANN), we train machine learners to classify music into known genres and compare the research with our clustering research.”
As Wayland’s representative at Texas Undergraduate Research Day, the international student from Australia received a $100 award plus a trip to Austin. Last month,Joffe’s research resulted in him winning the $500 Amir Moez Prize at the Texas Academy of Science Meeting at Angelo State University. The prize is awarded to the top mathematics and computer science oral presentation.
Undergraduate Research is an incredible opportunity for Wayland students to go beyond the classroom and make advances in their field of study while still at an undergraduate level. The university’s undergraduate research program has the distinction of being entirely student focused and student driven.
“Our students are involved in every stage of the research from asking the question, to designing the experiments, to collecting the data, to analyzing the results, to drawing conclusions, to presenting and defending their findings,” said Dr. Scott Franklin, Professor of Mathematics and Computer Science, who serves as Joffe’s research adviser. “It’s truly incredible how much our students accomplish under the direction of our exception faculty and the one-on-one approach taken.”
Joffe transferred to Wayland as a baseball player and a computer science major. Having significant coding and app development experience upon arrival, Franklin said he was eager to find ways beyond the computer science curriculum to apply his knowledge and skills.
“During his junior year, he began working with me to learn foundational concepts in machine learning and artificial intelligence,” Franklin said. “We considered a number of different projects of interest to Ilan and he came across a problem in using machine learners — a type of AI — to automatically classify music genres given the raw audio of a file.”
The research adviser described Joffe as “an exceptional student, whose programming skill and problem-solving ability surpasses most students. He has a strong work ethic and regularly pursues ideas and problem that are beyond the scope of our computer science curriculum.”
Copyright © 2022
