Music Genre Predictor based Classification of Audio Files with Low Level Feature of Frequency and Time Domain using Support Vector Machine Over K-Means Clustering Algorithm

Abstract

Main goal of the research is to employ Music genre prediction-based classification of audio files with low level feature of frequency domain and time domain using K-Means Clustering (K-Means) and Support Vector Machine (SVM). Materials and Methods: SVM and K-Means are implemented in this research work. Sample size is calculated using G power software and determined as 10 per group with pretest power 80%, threshold 0.05% and CI 95%. Result: SVM provides a higher of 95.35% compared to K-Means algorithm with 75.20% in predicting classification of Audio files with low level feature of frequency domain. There is a noteworthy difference between two groups with a significance value of 0.28 (p>0.05). Conclusion: NovelSupport Vector Machine algorithm predicts audio files with low level frequency better than K-Means algorithm.