Citas bibligráficas
Montenegro, R., (2018). Clasificación de especies forestales maderables de la Amazonía Peruana aplicando análisis Clúster con algoritmo Clara [Tesis, Universidad Nacional Agraria La Molina]. https://hdl.handle.net/20.500.12996/3759
Montenegro, R., Clasificación de especies forestales maderables de la Amazonía Peruana aplicando análisis Clúster con algoritmo Clara [Tesis]. : Universidad Nacional Agraria La Molina; 2018. https://hdl.handle.net/20.500.12996/3759
@mastersthesis{renati/243983,
title = "Clasificación de especies forestales maderables de la Amazonía Peruana aplicando análisis Clúster con algoritmo Clara",
author = "Montenegro Muro, Rolando Antonio",
publisher = "Universidad Nacional Agraria La Molina",
year = "2018"
}
The Amazon forest has a great variety of tree species, which is estimated at four thousand species. A few Amazonian species face extinction due to the high demand for their wood. To reduce the pressure on the demanded species and to promote the use of new species, it is necessary to know their technological characteristics, mainly, the physical and mechanical ones. Due to the large number of properties, it has been proposed to study the species with multivariate techniques, specifically through Cluster analysis. This with the purpose of grouping them according to the similarity they have in their physical and mechanical properties. Thus, it es possible to group little known species in the market with highly demanded species and suggest potential uses. For the present study, the CLARA algorithm (Clustering Large Applications) was used, which is suitable for large data sets. To select the optimal number of conglomerates, groups from two to ten were formed; then the average Silhouette width and the Dunn index were compared by group and the highest value was chosen. It was found that with an average Silhouette width of 0.339 the optimum number of clusters is two. The number of clusters indicated coincides with the results from the Dunn index, which reaches its highest value at 0.1264 with two clusters. The conglomerates had Guarea subridiflora (“requia de altura“) and Retrophyllum tospigliosii (“ulcumano“) as medoids. The first conglomerate was characterized by having high mechanical and physical properties, according to Aróstegui et al (1986). On the other hand, the “ulcumano“ medóide conglomerate was characterized by having low physical-mechanical properties, with the exception of the cleavage, which turned out to be medium.
Este ítem está sujeto a una licencia Creative Commons Licencia Creative Commons