Intelligent Music Composition

Automatic music composition is an enchanting field of research, inspiring both researchers and musicians. The implementation of systems that perform this task, incorporates an algorithmic part which makes decisions on which notes/sounds will be heard, when they will be heard, for how long and how loud. The intriguing part of automatic composition is the process of formulating such algorithms, in such a manner that the composed music is not too simple or too complex. Referring to the terms “simple” and “complex,” someone may find her/himself sort of words when speaking about music. Nevertheless, almost everyone can more easily express an opinion about the complexity of a certain music excerpt. This fact raises the question of how we may be sure that an automatic composition system composes too simple or too complex music. Additionally, how may we be able to compare the compositional capabilities of two such systems? Is there an objective criterion to characterize how “pleasant” is the music produced by such a machine? These questions among others have led the research community toward a direction to construct systems that have inherited some kind of intelligence, which allows them to construct music with a pleasant blend of simplicity and complexity. Specifically, these machines incorporate an intelligent algorithm, which is commonly able to create mathematical objects that may vary from numeric values to abstract string sequences. The underlying intelligence of these algorithms forges the aforementioned objects in a manner that they either produce present rich and complex structure or comply with certain target characteristics defined by humans. These objects could then simply be mapped to the sonic domain and produce subjectively pleasant music, or adjust the parameters of music composition systems and lead the composition process to certain stylistic and aesthetic directions. This chapter presents several theoretic and technical aspects of such systems and proposes their categorization according to their intelligence and the way that this