In the most recent literature, we may find many studies concerning biogeochemical features of ecosystems, communities' ecological structure, analysis of multi-temporal and multi-scales data sets coming from remote observations, atmospheric pollution modeling and forecasting in which multivariate procedures are applied . In all these cases multivariate statistical techniques have to be sequentially and recursively applied and tools able to compare the role of different variables in different correlation structures are necessary. In this study we present the formulation of two new indices based on the joined application of cluster analysis and principal component analysis. These indices are able to evaluate, quantitatively, a standardized weight for measured variables (descriptors) and object-observations or object-samples (objects), characterizing different correlation structures.
Multivariate indices for analysing correlation structures in environmental datasets
RAGOSTA, Maria;
2011-01-01
Abstract
In the most recent literature, we may find many studies concerning biogeochemical features of ecosystems, communities' ecological structure, analysis of multi-temporal and multi-scales data sets coming from remote observations, atmospheric pollution modeling and forecasting in which multivariate procedures are applied . In all these cases multivariate statistical techniques have to be sequentially and recursively applied and tools able to compare the role of different variables in different correlation structures are necessary. In this study we present the formulation of two new indices based on the joined application of cluster analysis and principal component analysis. These indices are able to evaluate, quantitatively, a standardized weight for measured variables (descriptors) and object-observations or object-samples (objects), characterizing different correlation structures.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
