Multivariate Datenanalyse

Henrion, René; Henrion, Günter

doi:10.1007/978-3-642-57792-5

Cited by 66 publications

(39 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Always when the term significant is used in the discussion, a P value lower than 0.05 is implied. An overview of the multivariate statistical methods applied is given in Dillon and Goldstein (1984) or Henrion and Henrion (1994). The factor analysis is an exploratory method to identify structures in multivariate databases by the reduction of the original multiple variables system to a simpler system with a reduced number of superimposed complex variables (factors).…”

Section: Methodsmentioning

confidence: 99%

Multivariate statistical approach to the temporal and spatial patterns of selected bioindicators observed in the North Sea during the years 1995?1997

Schmolke

Broeg

Zander

et al. 1999

Helgoland Marine Research

View full text Add to dashboard Cite

A comprehensive database, containing biological and chemical information, collected in the framework of the bilateral interdisciplinary MARS project ("biological indicators of natural and man-made changes in marine and coastal waters") during the years 1995-1997 in the coastal environment of the North Sea, was subjected to a multivariate statistical evaluation. The MARS project was designated to combine a variety of approaches and to develop a set of methods for the employment of biological indicators in pollution monitoring and environmental quality assessment. In total, nine ship cruises to four coastal sampling sites were conducted; 765 fish and 384 mussel samples were analysed for biological and chemical parameters. Additional information on the chemical background at the sampling sites was derived from sediment samples, collected at each of the four sampling sites. Based on the available chemical data in sediments and black mussel (Mytilus edulis) a pollution gradient between the selected sites, was established.The chemical body burden of flounder (Platichthys flesus) from these sites, though, did not reflect this gradient equally clear. In contrast, the biological information derived from measurements in fish samples displayed significant a regional as well as a temporal pattern. A multivariate bioindicator data matrix was evaluated employing a factor analysis model to identify relations between selected biological indicators, and to improve the understanding of a regional and temporal component in the parameter response. In a second approach, applying the kmeans algorithm on the data matrix, two significantly different clusters of samples, characterised by the current health status of the fish, were extracted. Using this classification a temporal, and in the second order, a less pronounced spatial effect was evident. In particular, during July 1996, a clear sign of deteriorating environmental conditions was extracted from the biological data matrix.

show abstract

Section: Methodsmentioning

confidence: 99%

Multivariate statistical approach to the temporal and spatial patterns of selected bioindicators observed in the North Sea during the years 1995?1997

Schmolke

Broeg

Zander

et al. 1999

Helgoland Marine Research

View full text Add to dashboard Cite

show abstract

“…Majority of parametric pattern recognition classification methods (LDA, SIMCA, PLS-DA) [4,10] works well when the classes present inner similarities and are well separated from each other (minimization of the ratio of within classes sum of squares and between classes sum of squares). Thus, when dealing with data exhibiting non-linear class structure, these methods are not a good choice; it should be more advisable to call on neural networks classification algorithms like the multilayer perceptron networks trained by error back propagation (MPN-BP) [11 -13].…”

Section: Supervised Pattern Recognition Using Neural Network Based Amentioning

confidence: 99%

“…Pattern recognition of brandies have been carried out by selecting organic trace compounds such as fusel oils [4,5]. In the present paper we introduce the use of metal profiles as possible chemical descriptors for the sake of brandies classification.…”

Section: Introductionmentioning

confidence: 99%

Differentiation of Spanish brandies according to their metal content

Cameán

Moreno

López-Artíguez

et al. 2001

Talanta

View full text Add to dashboard Cite

Eleven metals, namely, aluminium, calcium, cadmium, copper, iron, lead, magnesium, manganese, potassium, sodium and zinc were determined in twenty samples of Sherry brandies and twelve samples of Penedés brandies by applying atomic spectrometry techniques. Flame atomic absorption spectrometry was used for quantitating calcium, copper, iron, magnesium, manganese and zinc; atomic emission spectrometry to determine potassium and sodium; and graphite furnace atomic absorption spectrometry to analyse aluminium, cadmium and lead. A chemometric approach was followed to study the discrimination between brandies from Sherry or Penedés according to the metal profile.

show abstract

“…It is not a reflection of any special connection between maths and youth -a myth unsupported by the data 2,3 . As some mathematicians have long recognized 4 , this accident has been to mathematics' detriment. It reinforces biases within the discipline and in the public's attitudes about mathematicians' work, career pathways and intellectual and social values.…”

mentioning

confidence: 99%

The Fields Medal should return to its roots

Bárány

2018

Nature

View full text Add to dashboard Cite

Multivariate Datenanalyse

Cited by 66 publications

References 0 publications

Multivariate statistical approach to the temporal and spatial patterns of selected bioindicators observed in the North Sea during the years 1995?1997

Multivariate statistical approach to the temporal and spatial patterns of selected bioindicators observed in the North Sea during the years 1995?1997

Differentiation of Spanish brandies according to their metal content

The Fields Medal should return to its roots

Contact Info

Product

Resources

About