Servicios Personalizados
Revista
Articulo
Indicadores
- Citado por SciELO
- Accesos
Links relacionados
- Citado por Google
- Similares en SciELO
- Similares en Google
Compartir
Revista Ingenierías Universidad de Medellín
versión impresa ISSN 1692-3324versión On-line ISSN 2248-4094
Resumen
HURTADO, Jorge E. y BEDOYA R., Daniel. The relationships of Mercalli intensity to instrumental information as a pattern classification task. Rev. ing. univ. Medellin [online]. 2008, vol.7, n.13, pp.71-84. ISSN 1692-3324.
Despite the progress occurred in seismic instrumentation, the assessment of seismic vulnerability and damage qualitative indexes, such as that provided by Mercalli intensity is highly valuable and useful for practical purposes. In order to link the qualitative measures of earthquake action and its effects, statistical regression is commonly applied. in the paper, a different approach is adopted. It consisten in regarding the Mercalli intensity as a class rather that a numerical value. A modern statistical classification tool known as Support Vector Machine is used for classifying the la instrumental information in order to assess the corresponding Mercalli intensity. It is shown that the method gives satisfactory results with regardto the high uncertainties linked to such a qualitative seismic damage measure.
Palabras clave : Mercalli Intensity; earthquake damage; statistical learning; support vector machines; pattern recognition.