A wavelet, fourier, and PCA data analysis pipeline: Application to distinguishing mixtures of liquids

Münevver Köküer; Fionn Murtagh; Norman D. McMillan; Sven Riedel; Brian O'Rourke; Katie Beverly; Andy T. Augousti; Julian Mason

doi:10.1021/ci025601j

A wavelet, fourier, and PCA data analysis pipeline: Application to distinguishing mixtures of liquids

Münevver Köküer, Fionn Murtagh, Norman D. McMillan, Sven Riedel, Brian O'Rourke, Katie Beverly, Andy T. Augousti, Julian Mason

Research output: Contribution to journal › Article › peer-review

6 Citations (Scopus)

Abstract

Using a new optical engineering technique for the "fingerprinting" of beverages and other liquids, we study and evaluate a range of features. The features are based on resolution scale, invariant frequency information, entropy, and energy. They allow mixtures of beverages to be very precisely placed in principal component plots used for the data analysis. To show this we make use of data sets resulting from optical/near-infrared and ultrasound sensors. Our liquid "fingerprinting" is a relatively open analysis framework in order to cater for different practical applications, in particular, on one hand, discrimination and best fit between fingerprints, and, on the other hand, more exploratory and open-ended data mining.

Original language	English
Pages (from-to)	587-594
Number of pages	8
Journal	Journal of Chemical Information and Computer Sciences
Volume	43
Issue number	2
Early online date	23 Jan 2003
DOIs	https://doi.org/10.1021/ci025601j
Publication status	Published - 1 Mar 2003
Externally published	Yes

Access to Document

10.1021/ci025601jLicence: Unspecified

Cite this

@article{2b91a9b64072417bac987532cb265e92,

title = "A wavelet, fourier, and PCA data analysis pipeline: Application to distinguishing mixtures of liquids",

abstract = "Using a new optical engineering technique for the {"}fingerprinting{"} of beverages and other liquids, we study and evaluate a range of features. The features are based on resolution scale, invariant frequency information, entropy, and energy. They allow mixtures of beverages to be very precisely placed in principal component plots used for the data analysis. To show this we make use of data sets resulting from optical/near-infrared and ultrasound sensors. Our liquid {"}fingerprinting{"} is a relatively open analysis framework in order to cater for different practical applications, in particular, on one hand, discrimination and best fit between fingerprints, and, on the other hand, more exploratory and open-ended data mining.",

author = "M{\"u}nevver K{\"o}k{\"u}er and Fionn Murtagh and McMillan, {Norman D.} and Sven Riedel and Brian O'Rourke and Katie Beverly and Augousti, {Andy T.} and Julian Mason",

year = "2003",

month = mar,

day = "1",

doi = "10.1021/ci025601j",

language = "English",

volume = "43",

pages = "587--594",

journal = "Journal of Chemical Information and Modeling",

issn = "1549-9596",

publisher = "American Chemical Society",

number = "2",

}

TY - JOUR

T1 - A wavelet, fourier, and PCA data analysis pipeline

T2 - Application to distinguishing mixtures of liquids

AU - Köküer, Münevver

AU - Murtagh, Fionn

AU - McMillan, Norman D.

AU - Riedel, Sven

AU - O'Rourke, Brian

AU - Beverly, Katie

AU - Augousti, Andy T.

AU - Mason, Julian

PY - 2003/3/1

Y1 - 2003/3/1

N2 - Using a new optical engineering technique for the "fingerprinting" of beverages and other liquids, we study and evaluate a range of features. The features are based on resolution scale, invariant frequency information, entropy, and energy. They allow mixtures of beverages to be very precisely placed in principal component plots used for the data analysis. To show this we make use of data sets resulting from optical/near-infrared and ultrasound sensors. Our liquid "fingerprinting" is a relatively open analysis framework in order to cater for different practical applications, in particular, on one hand, discrimination and best fit between fingerprints, and, on the other hand, more exploratory and open-ended data mining.

AB - Using a new optical engineering technique for the "fingerprinting" of beverages and other liquids, we study and evaluate a range of features. The features are based on resolution scale, invariant frequency information, entropy, and energy. They allow mixtures of beverages to be very precisely placed in principal component plots used for the data analysis. To show this we make use of data sets resulting from optical/near-infrared and ultrasound sensors. Our liquid "fingerprinting" is a relatively open analysis framework in order to cater for different practical applications, in particular, on one hand, discrimination and best fit between fingerprints, and, on the other hand, more exploratory and open-ended data mining.

UR - http://www.scopus.com/inward/record.url?scp=18744433674&partnerID=8YFLogxK

U2 - 10.1021/ci025601j

DO - 10.1021/ci025601j

M3 - Article

AN - SCOPUS:18744433674

VL - 43

SP - 587

EP - 594

JO - Journal of Chemical Information and Modeling

JF - Journal of Chemical Information and Modeling

SN - 1549-9596

IS - 2

ER -

A wavelet, fourier, and PCA data analysis pipeline: Application to distinguishing mixtures of liquids

Abstract

Access to Document

Other files and links

Fingerprint

Cite this