The study of treatments effect is a very general issue, which encompasses several research fields from biomedicine, metabolomics to food processing and material science. The main framework is to assess if different treatments (factors), and which, affect the studied system and how. Using Analysis of Variance (ANOVA) has traditionally assessed these questions, however when several parameters are used to characterize the samples groups and, especially in the case of instrumental characterization, it has been demonstrated that coupling ANOVA with multivariate data analysis is far more efficient. In particular, ANOVA Simultaneous Component Analysis (ASCA) and ASCA combined with PARAFAC (PARAFASCA) have been proposed [1-2]. These methods, which are based on fitting a multivariate decomposition model on each ANOVA term, allows for easy interpretation of the variation induced by the different factors of the design due to the graphical representation in components space and projection/evaluation of residuals. In particular, ASCA describes each variance contribution with a PCA model, but a contribution d epending on crossed factors (interactions) may be described more parsimoniously by multiway models like parallel factor analysis (PARAFAC). In both cases designed data are described in a way that is both parsimonious and focused on the experimental question. The study of thermal treatments changes in moderate range of temperatures for spruce wood has been used as benchmark to illustrate the methodology. According to ISPM-15 standard, all wood materials to be shipped should be heat treated with specific time-temperature schedules. Thus, it is commercially important to have a fast method to estimate if a wood has been treated or not, to this aim Near Infrared Spectroscopy has been used as fingerprinting technique to evaluate if could be sensible to changes to wood chemical/physic structure induced by temperature and exposition time. Experiments have been planned according to a design taking into account three factors: temperature, treatment time and time occurred from treatment to the measurement.
Attenzione! Scheda prodotto non ancora validata dall'Ateneo
Dati e metadati della pubblicazione sono in fase di verifica da parte dell'Ateneo. In caso di errori o violazione dei diritti d'autore, contattare: firstname.lastname@example.org
|Data di pubblicazione:||2015|
|Titolo:||DIFFERENT APPROACHES TO THE ANALYSIS OF DESIGNED NIR FINGERPRINTING DATA|
|Appare nelle tipologie:||Abstract in Atti di Convegno|
File in questo prodotto:
I documenti presenti in Iris Unimore sono rilasciati con licenza Creative Commons Attribuzione - Non commerciale - Non opere derivate 3.0 Italia, salvo diversa indicazione.
In caso di violazione di copyright, contattare Supporto Iris