Premium
Characterisation of mineral waters by pattern recognition methods
Author(s) -
Caselli Maurizio,
De Giglio Angelo,
Mangone Annarosa,
Traini Angela
Publication year - 1998
Publication title -
journal of the science of food and agriculture
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.782
H-Index - 142
eISSN - 1097-0010
pISSN - 0022-5142
DOI - 10.1002/(sici)1097-0010(199804)76:4<533::aid-jsfa984>3.0.co;2-6
Subject(s) - principal component analysis , multivariate statistics , pattern recognition (psychology) , hierarchical clustering , data set , mineral , multivariate analysis , set (abstract data type) , data mining , mineralogy , statistics , mathematics , computer science , geology , artificial intelligence , chemistry , cluster analysis , organic chemistry , programming language
Eighty‐three samples of mineral water from four different wells in the same district were analysed for 23 parameters. Nineteen parameters were chosen for multivariate analysis. Principal components analysis provided a feature reduction to two or three dimensions without substantial loss of information. The data set is well separated into four clusters using hierarchical and non‐hierarchical methods; samples from different wells are generally assigned to different clusters. © 1998 SCI.