Support Vector Machine Applications in Terahertz Pulsed Signals Feature Sets

Yin, Xiaoxia; Ng, Brian W.-H.; Fischer, Bernd; Ferguson, Bradley S.; Abbott, Derek

doi:10.1109/jsen.2007.908243

Cited by 46 publications

(42 citation statements)

References 30 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This paper extends the formulation of a very important class of recently developed classifiers called Extreme Learning Machines (ELMs) to complex valued problems [15,16]. The motivation for the proposed extension stems from the fact that the real valued ELM has shown some of the lowest training errors among machine learning algorithms and in particular support vector machines classifiers (SVMs) [5,[20][21][22]. By extending ELMs to complex inputs, their applications domain can dramatically increase, encompassing all types of research associated to the study of the interaction of matter with waves, and in particular spectroscopy (acoustic, dielectric, optical, terahertz, infrared, electron-spin resonance, nuclear magnetic or paramagnetic resonance, etc.)…”

Section: Classification Methodologymentioning

confidence: 99%

“…Such investigations also have applications in security (e.g. fingerprinting of explosives and illicit drug detection [5,35,36]). Our goal is to demonstrate a generic feature extraction methodology that may be used across different THz data sets.…”

Section: Thz Pulse Measurements Of Six Different Types Of Powder Samplesmentioning

confidence: 99%

“…In previous works [5], we have shown that using signal processing techniques, it is possible to apodize [6] and de-noise the corresponding time-domain signatures [7] or spectra or alternatively model them using a range of modelling techniques adopted from the systems identification literature [8,9]. We have further demonstrated that such analysis may be directly performed in the time, frequency or even the wavelet domains [10][11][12][13].…”

Section: Introductionmentioning

confidence: 98%

See 2 more Smart Citations

Complex extreme learning machine applications in terahertz pulsed signals feature sets

Yin

Hadjiloucas

Zhang

2014

Computer Methods and Programs in Biomedicine

View full text Add to dashboard Cite

This paper presents a novel approach to the automatic classification of very large data sets composed of terahertz pulse transient signals, highlighting their potential use in biochemical, biomedical, pharmaceutical and security applications. Two different types of THz spectra are considered in the classification process. Firstly a binary classification study of poly-A and poly-C ribonucleic acid samples is performed. This is then contrasted with a difficult multi-class classification problem of spectra from six different powder samples that although have fairly indistinguishable features in the optical spectrum, they also possess a few discernable spectral features in the terahertz part of the spectrum. Classification is performed using a complex-valued extreme learning machine algorithm that takes into account features in both the amplitude as well as the phase of the recorded spectra. Classification speed and accuracy are contrasted with that achieved using a support vector machine classifier. The study systematically compares the classifier performance achieved after adopting different Gaussian kernels when separating amplitude and phase signatures. The two signatures are presented as feature vectors for both training and testing purposes. The study confirms the utility of complex-valued extreme learning machine algorithms for classification of the very large data sets generated with current terahertz imaging spectrometers. The classifier can take into consideration heterogeneous layers within an object as would be required within a tomographic setting and is sufficiently robust to detect patterns hidden inside noisy terahertz data sets. The proposed study opens up the opportunity for the establishment of complex-valued extreme learning machine algorithms as new chemometric tools that will assist the wider proliferation of terahertz sensing technology for chemical sensing, quality control, security screening and clinic diagnosis. Furthermore, the proposed algorithm should also be very useful in other applications requiring the classification of very large datasets.

show abstract

Section: Classification Methodologymentioning

confidence: 99%

Section: Thz Pulse Measurements Of Six Different Types Of Powder Samplesmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 98%

See 1 more Smart Citation

Complex extreme learning machine applications in terahertz pulsed signals feature sets

Yin

Hadjiloucas

Zhang

2014

Computer Methods and Programs in Biomedicine

View full text Add to dashboard Cite

show abstract

“…In this case, it is difficult to identify directly the samples by their characteristic spectral features. To realize identification, we can adopt pattern recognition methods to identify THz spectra of these samples [27][28]. Here, a PCA-WLDA pattern recognition method is introduced, in which PCA extracts the features spectroscopy and WLDA classification discrimination by using features spectroscopy is employed.…”

Section: Thz Spectrum Of Transgenic Cottonsmentioning

confidence: 99%

Method for identifying transgenic cottons based on terahertz spectra and WLDA

Liu

et al. 2015

Optik - International Journal for Light and Electron Optics

View full text Add to dashboard Cite

“…Consequently, the technological advances have been accompanied by much interest in possible applications of security [3][4][5], pharmaceutical [6][7], nondestructive testing (NDT) [8][9], drug inspection [10] and biological molecules testing [11][12][13][14].…”

Section: Introductionmentioning

confidence: 99%

Analysis of Terahertz Absorption Spectra Characteristic of 1-Naphthaleneacetic Acid and 4-Chlorophenoxyacetic Acid in Terahertz Region

Wang

et al. 2012

2012 International Conference on Industrial Control and Electronics Engineering

View full text Add to dashboard Cite

We firstly study terahertz time-domain spectroscopy (THz-TDS) characterization of two kinds of pesticides 1-naphthaleneacetic acid and 4-chlorophenoxyacetic acid. The absorption spectrum and refractive index of samples were measured. Theoretical investigation of the vibrational spectrum of pesticides was performed based on density function theory (DFT). We observed that experiment spectra are in rough qualitative agreement with these theory spectra. Its chemical composition was successfully identified through spectral analysis of the recorded terahertz waveform. These feature spectra measurements are helpful to form a database of pesticide.

show abstract

Support Vector Machine Applications in Terahertz Pulsed Signals Feature Sets

Cited by 46 publications

References 30 publications

Complex extreme learning machine applications in terahertz pulsed signals feature sets

Complex extreme learning machine applications in terahertz pulsed signals feature sets

Method for identifying transgenic cottons based on terahertz spectra and WLDA

Analysis of Terahertz Absorption Spectra Characteristic of 1-Naphthaleneacetic Acid and 4-Chlorophenoxyacetic Acid in Terahertz Region

Contact Info

Product

Resources

About