Efficient prediction of music genre using support vector machine and decision tree.

The study aims to evaluate the performance of the Music Genre Prediction System using a Novel Support Vector Machine and Decision Tree. The MARSYAS website provided the GTZAN dataset utilised in this investigation, which is comprised of one thousand.au-formatted music files and is intended for use in Music Information Retrieval. It is currently considered the gold standard dataset. Mel-frequency cepstralcoefficients (MFCC) are retrieved from music files and used to make predictions about the genre. All steps, from analysing data to training a model to testing it, take place within Jupyter. SPSS is used to analyse the mean accuracies of two algorithms side-by-side. Accuracy, precision, etc., numbers came out differently depending on the size of the input data set. SVM and Decision Tree are used for classification, with N equal to 20 for each of the two groups (proposed and comparative). The value of the pre-test achieved is 0.08. This research shows that compared to the Decision Tree method, Novel SVM is more effective. In the Research study, SVM with 81.10% accuracy is found more effective, efficient, and faster when compared to the Decision Tree with 59.88% accuracy. [ABSTRACT FROM AUTHOR]