Convolutional neural networks for classification and regression analysis of one-dimensional spectral data

Jernelv, Ine L.; Hjelme, Dag Roar; Matsuura, Yuji; Aksnes, Astrid

doi:10.48550/arxiv.2005.07530

Cited by 9 publications

(8 citation statements)

References 32 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For NIR data, the filter size of the pooling layers is set to two due to the smaller number of wavelengths of the Nir22 spectrometer. We use raw spectra as input data because CNNs generally show strong performance on unprocessed data, as outlined by Jernelv et al [ 33 ]. Furthermore, this approach saves computational effort.…”

Section: Methodsmentioning

confidence: 99%

Detection of Plastic Granules and Their Mixtures

Kulko

Pletl

Hanus³

et al. 2023

Sensors

View full text Add to dashboard Cite

Chemically pure plastic granulate is used as the starting material in the production of plastic parts. Extrusion machines rely on purity, otherwise resources are lost, and waste is produced. To avoid losses, the machines need to analyze the raw material. Spectroscopy in the visible and near-infrared range and machine learning can be used as analyzers. We present an approach using two spectrometers with a spectral range of 400–1700 nm and a fusion model comprising classification, regression, and validation to detect 25 materials and proportions of their binary mixtures. one dimensional convolutional neural network is used for classification and partial least squares regression for the estimation of proportions. The classification is validated by reconstructing the sample spectrum using the component spectra in linear least squares fitting. To save time and effort, the fusion model is trained on semi-empirical spectral data. The component spectra are acquired empirically and the binary mixture spectra are computed as linear combinations. The fusion model achieves very a high accuracy on visible and near-infrared spectral data. Even in a smaller spectral range from 400–1100 nm, the accuracy is high. The visible and near-infrared spectroscopy and the presented fusion model can be used as a concept for building an analyzer. Inexpensive silicon sensor-based spectrometers can be used.

show abstract

Section: Methodsmentioning

confidence: 99%

Detection of Plastic Granules and Their Mixtures

Kulko

Pletl

Hanus³

et al. 2023

Sensors

View full text Add to dashboard Cite

show abstract

“…A similar approach was followed later, using data obtained from the literature [41], as well as toenail samples [42]. Jernelv et al later employed convolutional neural networks for in vitro glucose detection measurements obtained from online datasets, including NIR and FTIR measurements [43]. However, no actual experimental measurements were conducted.…”

Section: Machine Learning Techniques For Glucose Detectionmentioning

confidence: 99%

A Single Wavelength Mid-Infrared Photoacoustic Spectroscopy for Noninvasive Glucose Detection Using Machine Learning

Aloraynan

Rassel

et al. 2022

Biosensors

View full text Add to dashboard Cite

According to the International Diabetes Federation, 530 million people worldwide have diabetes, with more than 6.7 million reported deaths in 2021. Monitoring blood glucose levels is essential for individuals with diabetes, and developing noninvasive monitors has been a long-standing aspiration in diabetes management. The ideal method for monitoring diabetes is to obtain the glucose concentration level with a fast, accurate, and pain-free measurement that does not require blood drawing or a surgical operation. Multiple noninvasive glucose detection techniques have been developed, including bio-impedance spectroscopy, electromagnetic sensing, and metabolic heat conformation. Nevertheless, reliability and consistency challenges were reported for these methods due to ambient temperature and environmental condition sensitivity. Among all the noninvasive glucose detection techniques, optical spectroscopy has rapidly advanced. A photoacoustic system has been developed using a single wavelength quantum cascade laser, lasing at a glucose fingerprint of 1080 cm−1 for noninvasive glucose monitoring. The system has been examined using artificial skin phantoms, covering the normal and hyperglycemia blood glucose ranges. The detection sensitivity of the system has been improved to ±25 mg/dL using a single wavelength for the entire range of blood glucose. Machine learning has been employed to detect glucose levels using photoacoustic spectroscopy in skin samples. Ensemble machine learning models have been developed to measure glucose concentration using classification techniques. The model has achieved a 90.4% prediction accuracy with 100% of the predicted data located in zones A and B of Clarke’s error grid analysis. This finding fulfills the US Food and Drug Administration requirements for glucose monitors.

show abstract

“…CNNs are dominant in various computer vision applications, such as facial recognition, target detection, image recognition, image annotation, image theme generation, image content generation and object annotation. However, CNNs are also used for data regression analysis (Jernelv et al, 2020). One-dimensional spectral data may be reshaped to two-dimensional arrays, and a 2D CNN model can be used for regression applications.…”

Section: Model Construction For Toc and Tss Calculationmentioning

confidence: 99%

Online water quality monitoring based on UV–Vis spectrometry and artificial neural networks in a river confluence near Sherfield-on-Loddon

Zhang

Wang

et al. 2022

Environ Monit Assess

View full text Add to dashboard Cite

Water quality monitoring is very important in agricultural catchments. UV–Vis spectrometry is widely used in place of traditional analytical methods because it is cost effective and fast and there is no chemical waste. In recent years, artificial neural networks have been extensively studied and used in various areas. In this study, we plan to simplify water quality monitoring with UV–Vis spectrometry and artificial neural networks. Samples were collected and immediately taken back to a laboratory for analysis. The absorption spectra of the water sample were acquired within a wavelength range from 200 to 800 nm. Convolutional neural network (CNN) and partial least squares (PLS) methods are used to calculate water parameters and obtain accurate results. The experimental results of this study show that both PLS and CNN methods may obtain an accurate result: linear correlation coefficient (R2) between predicted value and true values of TOC concentrations is 0.927 with PLS model and 0.953 with CNN model, R2 between predicted value and true values of TSS concentrations is 0.827 with PLS model and 0.915 with CNN model. CNN method may obtain a better linear correlation coefficient (R2) even with small number of samples and can be used for online water quality monitoring combined with UV–Vis spectrometry in agricultural catchment.

show abstract

Convolutional neural networks for classification and regression analysis of one-dimensional spectral data

Cited by 9 publications

References 32 publications

Detection of Plastic Granules and Their Mixtures

Detection of Plastic Granules and Their Mixtures

A Single Wavelength Mid-Infrared Photoacoustic Spectroscopy for Noninvasive Glucose Detection Using Machine Learning

Online water quality monitoring based on UV–Vis spectrometry and artificial neural networks in a river confluence near Sherfield-on-Loddon

Contact Info

Product

Resources

About