Topsis Based Ensemble Technique For Genre Classification

Abstract

Music Genre classification which comes under the area of Music Information Retrieval (MIR) has been an area of interest among researchers. A music genre is characterized by various features related to instrumentation, rhythmic structure, and form of members. To identify the genre of a given audio file has been a big challenge for the MIR community. This work describes an improved approach for classifying music into different genres. The proposed ensemble model is evaluated by using various parameters like accuracy, precision, recall and F1-score. Further, K-Fold cross-validation has been performed to check the consistency of the proposed ensemble model. To consider all the evaluation ranks the model by using the evaluation parameter. simultaneously, topsis a multicriteria decision analysis has been used. Musical genres are categorical labels created by humans to characterize pieces of music. A musical genre is characterized by the common characteristics shared by its members. These characteristics typically are related to the instrumentation, rhythmic structure, and harmonic content of the music. Genre hierarchies are commonly used to structure the large collections of music available on the Web. Currently, musical genre annotation is performed manually. Automatic musical genre classification can assist or replace the human user in this process and would be a valuable addition to music information retrieval systems. In addition, automatic musical genre classification provides a framework for developing and evaluating features for any type of content-based analysis of musical signals. In this paper, the automatic classification of audio signals into a hierarchy of musical genres is explored. More specifically, three feature sets for representing timbre texture, rhythmic content and pitch content are proposed. The performance and relative importance of the proposed features are investigated by training statistical pattern recognition classifiers using real world audio collections. Both the whole file and real-time frame-based classification schemes are described. We examine the performance of different classifiers on different audio feature sets to determine the genre of a given music piece. For each classifier, we also evaluate performances of feature set obtained by dimensionality reduction methods. We applied various machine learning models on data obtained from simulation to predict the rate and compare their performances with each other to find the best machine learning model. To check the robustness of the best model, we used k-fold cross-validation.