E-Locus - Institutional Repository of the University of Crete - Ταυτοποίηση και μελέτη γνησιότητας μελιού, ελαιολάδου και κρασιού με την χρήση οπτικών φασματοσκοπικών τεχνικών σε συνδυασμό με μεθόδους μηχανικής μάθησης

Home Ταυτοποίηση και μελέτη γνησιότητας μελιού, ελαιολάδου και κρασιού με την χρήση οπτικών φασματοσκοπικών τεχνικών σε συνδυασμό με μεθόδους μηχανικής μάθησης

Results - Details

[Add to Basket]

Identifier

000428370

Title

Ταυτοποίηση και μελέτη γνησιότητας μελιού, ελαιολάδου και κρασιού με την χρήση οπτικών φασματοσκοπικών τεχνικών σε συνδυασμό με μεθόδους μηχανικής μάθησης

Alternative Title

Identification and study of authenticity of honey, olive oil and wine with the use of optical spectroscopic techniques with the combination of machine learning methods

Author

Ορφανάκης, Εμμανουήλ Σ.

Thesis advisor

Χανιωτάκης, Νικόλαος

Reviewer

Βελεγράκης, Μιχάλης
Σαμαρτζής, Πέτρος

Abstract

The identification and study of authenticity of agri-food products is a need that focuses on the interest of both the general public and producers. High valueadded agri-food products such as wine, olive oil and honey are at the heart of this field as their geographical and botanical origin can be falsified. Classic analytical techniques that are applied to ensure the products quality and origin are accurate but time consuming. This study uses spectroscopic absorption, fluorescence and Raman techniques to systematically record optical spectra as characteristic ''fingerprints'' reflecting the chemical composition of agri-food products. The optical data were used in combination with machine learning models to determine the botanical origin of wine from different wineries and varieties, the botanical origin of honey derived from flowers, thyme and pine trees and the geographical origin of olive oil manufactured in Peloponnese, Crete and Lesvos. Models built by Principal Component Analysis (PCA), partial least squares (PLS), partial least squares discriminant analysis (PLS-DA) methodologies, where capable of distinguishing the botanical origin of the wine and honey samples and the geographical origin of the olive oil samples with high success rates, demonstrating the applicability of spectroscopic techniques in this field.

Language

Greek

Subject

Absorption spectroscopy

Botanical origin

Fluorescence spectroscopy

Geographical origin

Honey

Olive oil

PCA

PLS-DA

Partial least squares discriminant

Principles component analysis

Raman spectroscopy

Wine

Ανάλυση διακριτότητας μερικών ελαχίστων τετραγώνων

Ανάλυση κύριων συνιστωσών

Βοτανική προέλευση

Γεωγραφική προέλευση

Ελαιόλαδο

Κρασί

Μέλι

Φασματοσκοπία Raman

Issue date

2020-03-27

Collection

School/Department--School of Sciences and Engineering--Department of Chemistry--Post-graduate theses

Type of Work--Post-graduate theses

Views

561

Digital Documents
	Download document