Multiclass Support Vector Machines With Example-Dependent Costs Applied to Plankton Biomass Estimation

In many applications, the mistakes made by an automatic classifier are not equal, they have different costs. These problems may be solved using a cost-sensitive learning approach. The main idea is not to minimize the number of errors, but the total cost produced by such mistakes. This brief presents...

Descripción completa

Detalles Bibliográficos
Autores: González, Pablo, Álvarez, Eva, Barranquero, José, González-Quirós, Rafael, Nogueira, Enrique, López-Urrutia-Lorente, Ángel, Coz, Juan José
Tipo de recurso: artículo
Estado:Versión aceptada para publicación
Fecha de publicación:2013
País:España
Institución:Consejo Superior de Investigaciones Científicas (CSIC)
Repositorio:DIGITAL.CSIC. Repositorio Institucional del CSIC
OAI Identifier:oai:digital.csic.es:10261/319169
Acceso en línea:http://hdl.handle.net/10261/319169
Access Level:acceso abierto
Palabra clave:Centro Oceanográfico de Gijón
Medio Marino
Multiclass Support Vector Machines
plankton
Descripción
Sumario:In many applications, the mistakes made by an automatic classifier are not equal, they have different costs. These problems may be solved using a cost-sensitive learning approach. The main idea is not to minimize the number of errors, but the total cost produced by such mistakes. This brief presents a new multiclass cost-sensitive algorithm, in which each example has attached its corresponding misclassification cost. Our proposal is theoretically well-founded and is designed to optimize cost-sensitive loss functions. This research was motivated by a real-world problem, the biomass estimation of several plankton taxonomic groups. In this particular application, our method improves the performance of traditional multiclass classification approaches that optimize the accuracy.