Music as a formal language. Comparative analysis of audio content recording techniques using symbolic and subsymbolic representation for processing by machine learning methods
| Authors | |
|---|---|
| Year of publication | 2022 |
| Citation | |
| Description | Machine learning finds its use in solving tasks where there is no domain-specific knowledge. Systems that can extract essential regularities from the training set can be found in the sphere of unsupervised learning. These or similar algorithms can be useful tools for analyzing and modeling musical styles from corpora where we do not have sufficient domain-specific knowledge to create an independent domain. The generation of terminal elements of syntactic structure is essential for corpus analysis, hence the subsequent generation of new musical material. In the case of musical composition, terminals may take the form of chords, harmonic changes, melodic fragments, rhythmic figures or even playing techniques of specific instruments. Although this approach evokes a cliché, music and language have more in common than it may seem at first sight, at least in terms of computer technology. |
| Related projects: |