Terahertz Signal Classification Based on Geometric Algebra

This paper presents an approach to classification of substances based on their terahertz spectra. We use geometric algebra to provide a concise mathematical means for attacking the classification problem in a coordinate-free form. For the first time, this allows us to perform classification independ...

Full description

Saved in:
Bibliographic Details
Published inIEEE transactions on terahertz science and technology Vol. 6; no. 6; pp. 793 - 802
Main Authors Shengling Zhou, Valchev, Dimitar G., Dinovitser, Alex, Chappell, James M., Iqbal, Azhar, Ng, Brian Wai-Him, Kee, Tak W., Abbott, Derek
Format Journal Article
LanguageEnglish
Published Piscataway IEEE 01.11.2016
The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
Subjects
Algebra
Australia
Classification
Dispersion
Free form
geometric algebra
Microwaves
multivectors
Refractive index
Seismology
Signal classification
Signal to noise ratio
Source code
spectroscopy
Support vector machines
terahertz (THz)
Terahertz frequencies
Transfer functions
Online AccessGet full text

Cover

More Information
Summary:This paper presents an approach to classification of substances based on their terahertz spectra. We use geometric algebra to provide a concise mathematical means for attacking the classification problem in a coordinate-free form. For the first time, this allows us to perform classification independently of dispersion and, hence, independently of the transmission path length through the sample. Finally, we validate the approach with experimental data. In principle, the coordinate-free transformation can be extended to all types of pulsed signals, such as pulsed microwaves or even acoustic signals in the field of seismology. Our source code for classification based on geometric algebra is publicly available at: https://github.com/swuzhousl/Shengling-zhou/blob/geometricalgebra-classifier/GAclassifier/.
ISSN:2156-342X
2156-3446
DOI:10.1109/TTHZ.2016.2610759