Comparative Analysis of Machine Learning Approaches for Drum Set Component Classification in Music Technology

The drum set is the fundamental instrument of many musical genres. Although it is considered a whole musical instrument, it comprises different sizes and types of components. These components play a critical role in determining the rhythm, tempo, and form of the music as well as the musical style. In this study, we investigate using different machine learning techniques for the automatic classification of drum set components. The dataset was constructed using spectral and temporal features extracted from audio recordings of various drum set components. For feature selection and optimization, the Recursive Feature Elimination with Cross-Validation (RFECV) method combined with Gradient Boosting (GB) played a critical role. The resulting feature set was used to train classification models with Support Vector Machine (SVM), Random Forest (RF), K-Nearest Neighbor (KNN), and Decision Tree (DT) algorithms. The RF algorithm achieved the highest accuracy rate of 97.1% in classifying drum components.