Classifying Music Performance: Using k-Nearest Neighbour Classification - Couverture souple

Soh, Chew May Joyce; Tischer, Peter; Shetty, Pravin

 
9783639245424: Classifying Music Performance: Using k-Nearest Neighbour Classification
Couverture souple
ISBN 10 : 3639245423 ISBN 13 : 9783639245424
Editeur : VDM Verlag Dr. Müller, 2010

Synopsis

A performance is the execution of a sequence of actions. In a musical performance, some of the actions are dictated by the score. Other decisions made by the performer give the performance its character and individuality. Therefore, this project aims to identify these kinds of decisions and to extract information about these in as features from the performance. A computer program can be used to analyze the features of a performance and to classify whether a performance is consistent with a given performance profile. We propose and investigate two approaches for determining the underlying beat of a performance. A general structure for k-nearest neighbor (k-NN) classifier was developed. This involved consideration of 4 main issues: the choice of k, the choice of distance function, the way information from the nearest neighbors is used to produce the classification and how a k-NN classifier can indicate what confidence might be placed in the classification it returns. A general review of k-NN classification schemes with respect to these 4 main issues was undertaken, as was a design for software implementing a general k-NN classification scheme.

Les informations fournies dans la section « Synopsis » peuvent faire référence à une autre édition de ce titre.

Présentation de l'éditeur

A performance is the execution of a sequence of actions. In a musical performance, some of the actions are dictated by the score. Other decisions made by the performer give the performance its character and individuality. Therefore, this project aims to identify these kinds of decisions and to extract information about these in as features from the performance. A computer program can be used to analyze the features of a performance and to classify whether a performance is consistent with a given performance profile. We propose and investigate two approaches for determining the underlying beat of a performance. A general structure for k-nearest neighbor (k-NN) classifier was developed. This involved consideration of 4 main issues: the choice of k, the choice of distance function, the way information from the nearest neighbors is used to produce the classification and how a k-NN classifier can indicate what confidence might be placed in the classification it returns. A general review of k-NN classification schemes with respect to these 4 main issues was undertaken, as was a design for software implementing a general k-NN classification scheme.

Les informations fournies dans la section « A propos du livre » peuvent faire référence à une autre édition de ce titre.

Éditeur
VDM Verlag Dr. Müller
Date d'édition
2010
Langue
anglais
ISBN 10 
3639245423
ISBN 13 
9783639245424
Reliure
Broché
Nombre de pages
164
Coordonnées du fabricant
non disponible
Personne responsable
non disponible