An investigation of the fusion path of multiple singing styles in American vocal singing by incorporating Markov chain prediction model

Abstract

In this paper, by analyzing the four American vocal singing styles, the vocal semantics beyond the semantics of the four vocal singing styles are constructed by using the common knowledge map, and the heavy semantic fusion mechanism is established based on the pre-training model of the knowledge map to obtain the contextual semantic features, and the relational classification model MSF-RC is realized. To predict the degree of fusion of the three vocal singing styles, the classical gray theory and To predict the degree of fusion of the three vocal singing styles, the Markov chain prediction is completed for the residual numerical sequences with large volatility. After testing, the accuracy of this algorithm is 0.88, the recall is 0.92, the F-Score is 0.88, and the MAPE value between the actual values of American singing and ethnic singing fusion and the prediction results corrected by the fusion Markov chain model is 1%, which has high prediction accuracy.

Keywords

• knowledge graph
• semantic fusion
• relational classification
• gray theory
• markov chain
• american vocal singing
• 97b20