Mood of Song Detection Using Mel Frequency Cepstral Coefficient and Convolutional Neural Network with Tuning Hyperparameter
Wildan Budiawan Zulfikar, Yana Aditia Gerhana, Aulia Yasmin Putri Almi, Dian Sa’adillah Maylawati, Muhammad Insan Al-Amin · 2023
Music is divided into various types, one of which is divided according to the character of the music itself. Human have a tendency to listen to music based on mood because music have a close bond with the condition of human heart. This study tries to identify the types of musical moods for music classification using the Mel-frequency cepstral coefficients (MFCC) and Convolutional Neural Networks (CNN) Algorithm. This method is one of the methods that has been widely used in the field of Speech Technology, where this method aims to perform feature extraction and sound classification. The RMSprop optimizer with Uniform kernel as a hyperparameter tuning model produced the study’s most ideal outcomes, according to test findings utilizing the K-Fold cross validation technique. As a consequence, average values for accuracy, precision, recall, and F1-Score were 79.55%, 85%, 79%, and 81% respectively.