About this document
Hit Song Prediction Through Collaboration by yokorniienko.frgf24 is a document available to read on EtoBox.
The thesis presents a multimodal approach to analyze and predict hit songs by fusing features from music, artist, and album. It identifies collaboration profiles within a network of successful artists and examines the causal relationship between these profiles and artist popularity using time series and Granger causality tests. The prediction problem is modeled as two tasks: classification and placement, demonstrating the effectiveness of the proposed method through extensive empirical studies.
- Author
- yokorniienko.frgf24
- Language
- EN