A Review of Machine Learning for Near-Infrared Spectroscopy

Abstract

The analysis of infrared spectroscopy of substances is a non-invasive measurement technique that can be used in analytics. Although the main objective of this study is to provide a review of machine learning (ML) algorithms that have been reported for analyzing near-infrared (NIR) spectroscopy from traditional machine learning methods to deep network architectures, we also provide different NIR measurement modes, instruments, signal preprocessing methods, etc. Firstly, four different measurement modes available in NIR are reviewed, different types of NIR instruments are compared, and a summary of NIR data analysis methods is provided. Secondly, the public NIR spectroscopy datasets are briefly discussed, with links provided. Thirdly, the widely used data preprocessing and feature selection algorithms that have been reported for NIR spectroscopy are presented. Then, the majority of the traditional machine learning methods and deep network architectures that are commonly employed are covered. Finally, we conclude that developing the integration of a variety of machine learning algorithms in an efficient and lightweight manner is a significant future research direction.

The diagnostic value of pulmonary near‐infrared spectroscopy in the early distinction of neonatal pneumonia from transient tachypnea of the newborn

“…When a substance is exposed to near‐infrared light, infrared active molecular bonds interact with the light to produce NIRS absorption. NIRS is highly suitable for analyzing a wide range of samples, including liquids, solids and gases 16 . Furthermore, NIRS can be applied to both organic and inorganic materials 16 .…”

“…16 Furthermore, NIRS can be applied to both organic and inorganic materials. 16 Although NIRS is mostly used in neonatology to obtain real-time tissue oxygenation in solid organs (brain, kidney, mesenteric tissue, muscle, etc. ), [7][8][9] it can also be used in partially air-filled organs such as the neonatal lung.…”

“…NIRS is highly suitable for analyzing a wide range of samples, including liquids, solids and gases 16 . Furthermore, NIRS can be applied to both organic and inorganic materials 16 . Although NIRS is mostly used in neonatology to obtain real‐time tissue oxygenation in solid organs (brain, kidney, mesenteric tissue, muscle, etc.…”

Methodology adjusting for least squares regression slope in the application of multiplicative scatter correction to near‐infrared spectra of forage feed samples

“…9,10 Aiming at decreasing the impact of the baseline shifts and mitigating the effects of multicollinearity, spectral pre-treatments or transformations are commonly applied prior to calibration modelling. [11][12][13] The most extensively used spectral pre-treatments are multiplicative scatter (or signal) correction (MSC), 8 standard normal variate (SNV) 14 and detrending to remove the linear and quadratic trends of each spectrum. 14 Spectral scales will be specific to the transformations applied, 10 either alone or in combination with derivatives and/or smoothing.…”

“…The consequences of these effects are that NIR spectra often exhibit baseline shifts and that NIR calibration sets will entail some degree of multicollinearity 9,10 . Aiming at decreasing the impact of the baseline shifts and mitigating the effects of multicollinearity, spectral pre‐treatments or transformations are commonly applied prior to calibration modelling 11–13 . The most extensively used spectral pre‐treatments are multiplicative scatter (or signal) correction (MSC), 8 standard normal variate (SNV) 14 and detrending to remove the linear and quadratic trends of each spectrum 14 .…”