Electronic Dance Music Classification Based on Machine Learning Methods
Zihan Dai, Xuyan Huang · 2022 International Conference on Electronics and Devices, Computational Science (ICEDCS) · 2022
Music classification is an emerging area that much research focusses on in the past studies. However, electronic dance music (EDM) that is a famous type of music is rarely studied. In this study, several machine learning algorithms e.g. support vector machine (SVM), random forest (RF) and artificial neural network (ANN) were employed to achieve the classification for EDM. To be more specific, the data is collected from numerous playlists on Spotify. Then the data was preprocessed e.g. normalization firstly and then it was passed into different models for training and testing. The developed neural network consists of four fully connected layers. The accuracy based on the training and testing dataset and the confusion matrix were employed to find which model can achieve better result. The experimental results indicated that random forest could achieve the accuracy of 0.83 in the testing test, which is higher than other compared machine learning methods.