Music statistics: uncertain logistic regression models with applications in analyzing music.

In the realm of data analysis, traditional statistical methods often struggle when faced with ambiguity and uncertainty inherent in real world data. Uncerainty theory, developed to better model and interpret such data, offers a promising alternative to conventional techniques. In this paper, we establish logistic regression models to initiate music statistics based on uncertainty theory. In particular, we will classify the music into different types named Baroque, Classical, Romantic, and Impressionism based on four characteristics: harmonic complexity, rhythmic complexity, texture complexity, and formal structure, with the help of the uncertain logistic models proposed. This theoretical framework for music classification provides a nuanced understanding of how music is interpreted under conditions of ambiguity and variability. Compared with the probabilistic counterpart, our approach highlights the versatility of uncertainty theory and provides researchers one much more feasible method to analyze the often-subjective nature of music reception, as well as broadening the potential applications of uncertainty theory. [ABSTRACT FROM AUTHOR]