Cristina Miranda
A proposal for robust estimation of the extremal index
Many practical problems deal with extreme values above fixed levels and occurring in clusters of exceedances. The dependence among clusters is characterized by the extremal index of the process, which can be interpreted as the reciprocal of the clusters size mean. Several estimators of the extremal index have been proposed. They essentially differ in the identification of the groups. One of the most used proposals is known as the blocks estimator. Under specific local dependence and stationarity conditions the limit distribution of the number of exceedances follows a compound Poisson process, where clusters occurrences are modelled by a Poisson distribution and clusters dimensions are determined by the process multiplicities. Since clusters dimension is generally reduced and can have a discrete asymmetric distribution, a robust version of the estimator should be based on a robust estimator of the mean clusters dimension, taking into account those distributional properties. Assuming the Poisson distribution of the number of exceedances per cluster, the extremal index is estimated considering a robust Poisson regression estimate of the mean dimension of the clusters. The performance of the method is evaluated through the comparison of the results with those obtained when the observations are generated by a contaminated distribution.
Helena Alvelos
Statistical methods in Sensory Analysis – an overview
Tasting Panels are used in Sensory Analysis in order to evaluate products according to the way they are perceived by human senses. The main task of a professional taster is, therefore, to assess the sensorial characteristics of products: aspect, odour, consistency and texture, taste and noise. This type of panels are widely used in the food industry, namely, wine, olive oil, chocolate, among others, as well as in other types of industry, for example cosmetics and perfumes.
The tasting process usually generates a large amount of data that is used in decisions about the products and can also be used to evaluate the tasters. Evaluating the individual performance of the tasters is essential, so the assessments’ results are as reliable as possible.
In this context, statistical methods are commonly used so that tasters control can be carried out and tasting panels’ decisions can be supported by the available information.
In this talk, some of the most used statistical methods in tasters’ performance evaluation are briefly presented.
Analysis of Variance (ANOVA) has been one of the most often employed tools to analyze the tasters’ and panel performance. The panel performance is evaluated by two-way or three-way ANOVA. The method is based on modelling samples, tasters and their interactions in two-way ANOVA, or samples, tasters, repetitions and their interactions in three-way ANOVA.
Principal Component Analysis (PCA) and Cluster Analysis are some of the most used multivariate methods for studying panel performance. PCA is generally used to study the level of agreement within the panel and takes into account all the samples’ evaluations of each taster regarding the different characteristics under assessment. Tasters with similar behaviors will naturally be closer to each other in the resulting components.
Cluster analysis can be used to group judges based on the similarity of their sensory profiles across a set of samples. This method is usually applied to identify tasters that are not in accordance with the panel as a whole.
There are several other techniques and different approaches in the context of assessors’ and panel performance measurement that have been applied to different industries. Nonetheless, there is still a great deal of potential for the development of the use of Statistics in Sensory Analysis.
This is joint work with Ana Raquel Xambre, Leonor Teixeira and Ana Luísa Ramos.
The tasting process usually generates a large amount of data that is used in decisions about the products and can also be used to evaluate the tasters. Evaluating the individual performance of the tasters is essential, so the assessments’ results are as reliable as possible.
In this context, statistical methods are commonly used so that tasters control can be carried out and tasting panels’ decisions can be supported by the available information.
In this talk, some of the most used statistical methods in tasters’ performance evaluation are briefly presented.
Analysis of Variance (ANOVA) has been one of the most often employed tools to analyze the tasters’ and panel performance. The panel performance is evaluated by two-way or three-way ANOVA. The method is based on modelling samples, tasters and their interactions in two-way ANOVA, or samples, tasters, repetitions and their interactions in three-way ANOVA.
Principal Component Analysis (PCA) and Cluster Analysis are some of the most used multivariate methods for studying panel performance. PCA is generally used to study the level of agreement within the panel and takes into account all the samples’ evaluations of each taster regarding the different characteristics under assessment. Tasters with similar behaviors will naturally be closer to each other in the resulting components.
Cluster analysis can be used to group judges based on the similarity of their sensory profiles across a set of samples. This method is usually applied to identify tasters that are not in accordance with the panel as a whole.
There are several other techniques and different approaches in the context of assessors’ and panel performance measurement that have been applied to different industries. Nonetheless, there is still a great deal of potential for the development of the use of Statistics in Sensory Analysis.
This is joint work with Ana Raquel Xambre, Leonor Teixeira and Ana Luísa Ramos.