**audio **In Music Information Retrieval: analysis of music recordings

**background corpus **A collection of musical pieces on which training of computational model (e.g., models for expectancy) takes place.

**binary **A variable with two outcomes, e.g., true vs. false, or occurrence vs. non-occurrence.

**classification** The procedure of sorting items into categories, e.g. to assign melodies to tune families.

**clustering** The procedure of grouping items which are similar, e.g. to group melodies without prior categories such as tune families.

**dependent variable** In a statistical model, observations that are predicted by the model.

**diachronic** A diachronic analysis investigates change of cultural items over time.

**features** In machine learning, another word for variables of models.

**independent variable** In a statistical model, the property that is used to predict observations.

**inter-onset interval** The space between two note onsets.

**Music Information Retrieval** A research field which strives to obtain information from music recordings or scores through computational analysis.

**node** A connected element in a network model.

**neural network** A network model which is inspired by connections of neurons in the human brain, consisting of layers of nodes.

**onset** The start of a note, measured from the start of the piece.

**pitch chroma** Pitch chroma describes the different categories of pitches within an octave.

**recursion** A structure which is described in terms of repetitions of itself. See also recursion.

**regression** A statistical model which infers whether the development of a given property (independent variable) is statistically related to a given observation (dependent variable).

**symbolic **In Music Information Retrieval: analysis of music notation

**synchronic** A synchronic analysis investigates change of cultural items from the same time.

**viewpoint ** A musical aspect being studied, such as the pitch or duration of a note.