Machine learning concept in de-spiking process for nuclear resonant vibrational spectra – Automation using no external parameter

Wang, Jessie; Li, Lei; Wang, Hongxin

doi:10.1016/j.vibspec.2022.103352

Cited by 6 publications

(15 citation statements)

References 32 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In fact, cosmic rays do not necessarily appear with these patterns and these patterns may also be shared by other Raman peaks, leading to incomplete clearance of the cosmic rays (under a loose standard) or false deletion/reduction of some Raman peaks (under a strict standard). [108,263,264] Therefore, CNNs have been introduced to automatically extract the characteristics of the cosmic rays with least human intervention. Since the cosmic rays do not appear to disturb every spectrum, a high rate of spikes should be specifically added to create an adequate training set.…”

Section: Ai For Raman Instrumentations and Spectral Preprocessingmentioning

confidence: 99%

Artificial Intelligence for Surface‐Enhanced Raman Spectroscopy

Bi,

Lin,

Chen

et al. 2023

Small Methods

View full text Add to dashboard Cite

Surface‐enhanced Raman spectroscopy (SERS), well acknowledged as a fingerprinting and sensitive analytical technique, has exerted high applicational value in a broad range of fields including biomedicine, environmental protection, food safety among the others. In the endless pursuit of ever‐sensitive, robust, and comprehensive sensing and imaging, advancements keep emerging in the whole pipeline of SERS, from the design of SERS substrates and reporter molecules, synthetic route planning, instrument refinement, to data preprocessing and analysis methods. Artificial intelligence (AI), which is created to imitate and eventually exceed human behaviors, has exhibited its power in learning high‐level representations and recognizing complicated patterns with exceptional automaticity. Therefore, facing up with the intertwining influential factors and explosive data size, AI has been increasingly leveraged in all the above‐mentioned aspects in SERS, presenting elite efficiency in accelerating systematic optimization and deepening understanding about the fundamental physics and spectral data, which far transcends human labors and conventional computations. In this review, the recent progresses in SERS are summarized through the integration of AI, and new insights of the challenges and perspectives are provided in aim to better gear SERS toward the fast track.

show abstract

Section: Ai For Raman Instrumentations and Spectral Preprocessingmentioning

confidence: 99%

Artificial Intelligence for Surface‐Enhanced Raman Spectroscopy

Bi,

Lin,

Chen

et al. 2023

Small Methods

View full text Add to dashboard Cite

show abstract

“…NRVS spectra were recorded using a published procedure [28,48,49,52] at SPring-8 beamlines BL09XU in Japan. A high-heat load monochromator (HHLM) produced 14.4 keV radiation with ~1.0 eV energy resolution, and a high energy resolution monochromator (HRM) (with crystals of Ge(422) and Si(975)x2) [58,59] subsequently produced 14.4 keV radiation with 0.8 meV energy resolution, suitable for measuring vibrations.…”

Section: Nrvs Measurementsmentioning

confidence: 99%

“…In comparison with the conventional infrared absorption spectroscopy (IR) [50] (Figure 1e), it measures vibrations indirectly but has several distinct advantages. The most prominent advantages include but are not limited to being isotope (e.g., 57 Fe) specific for studying complicated systems [28,48,49,51,52], having an almost zero background [28,48], and being able to obtain a theoretically calculable partial vibrational density of state (PVDOS) [28,[45][46][47]53]. These advantages make it a better method in comparison with the laboratory-based IR [28,50], Raman [28] and laser-induced fluorescence (LIF) spectroscopies [54][55][56] as well as SR-based inelastic X-ray scattering (IXS) [28,57].…”

Section: Introductionmentioning

confidence: 99%

“…These advantages make it a better method in comparison with the laboratory-based IR [28,50], Raman [28] and laser-induced fluorescence (LIF) spectroscopies [54][55][56] as well as SR-based inelastic X-ray scattering (IXS) [28,57]. This modern spectroscopy became available in the mid 1990s, due to the development of the third-generation SR sources which provide the strong X-ray pulses with specific timing structure, advanced X-ray optics which lead to a monochromator with an 1 meV energy resolution to measure vibrations, and modern detectors which extract a weak nuclear scattering signal from the huge electronic scattering counts in the time domain, and in turn, it has pushed a great advancement in physics, chemistry, biochemistry and materials science, etc., for the past 28 years [45][46][47][48][49]52]. In chemistry and bioinorganic chemistry, for example, this technique has uncovered Fe-S/P/Cl, Fe-CO/CN/NO and Fe-H/D vibrational modes inside various inorganic complexes as well as dilute iron enzymes [28], and has thus become an excellent pinpointing tool to study iron-specific electronic and structural properties, including iron oxidation state(s) and coordination environments.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

X-ray and Nuclear Spectroscopies to Reveal the Element-Specific Oxidation States and Electronic Spin States for Nanoparticulated Manganese Cyanidoferrates and Analogs

Wang,

Huang,

Young

et al. 2023

Physchem

Self Cite

View full text Add to dashboard Cite

In this publication, the potential non-gadolinium magnetic resonant imaging agent—nanoparticulate K2Mn[Fe(CN)6]—its comparison sample KFe[Co(CN)6], as well as their reference samples were measured and analyzed using Mn, Co and Fe L-edge X-ray absorption spectroscopy (L XAS). From the information obtained, we conclude that K2Mn[Fe (CN)6] has a high spin (hs)-Mn(II) and a low spin (ls)-Fe(II), while KFe[Co(CN)6] has an hs-Fe(II) and an ls-Co(III). In these Prussian blue (PB) analog structures, the L XAS analysis also led to the conclusion that the hs-Mn(II) in K2Mn[Fe(CN)6] or the hs-Fe(II) in KFe[Co(CN)6] bonds to the N in the [M(CN)6]4−/3− ions (where M = Fe(II) or Co(III)), while the ls-Fe(II) in K2Mn[Fe(CN)6] or the ls-Co(III) in KFe[Co(CN)6] bonds to the C in the [M(CN)6]4−/3− ion, suggesting the complexed metalloligand [Mn(II) or Fe(II)] occupies the N-bound site in PB. Then, nuclear resonant vibrational spectroscopy (NRVS) was used to confirm the results from the L XAS measurements: the Mn(II), Eu(III), Gd(III), Fe(II) cations complexed by [M(CN)6]n−-metalloligand all take the N-bound site in PB-like structures. Our NRVS studies also prove that iron in the K2Mn[Fe(CN)6] compound has a 2+ oxidation state and is surrounded by the C donor atoms in the [M(CN)6]n− ions.

show abstract

“…Machine learning (ML) techniques are widely used to extract knowledge and data pattern from statistics and automate data analysis (Wang et al 2022). They have the capacity to build classification, prediction and/or regression models for various applications -mainly stock markets.…”

Section: Introductionmentioning

confidence: 99%

A federated learning-enabled predictive analysis to forecast stock market trends

Ardakani

Cai

et al. 2023

J Ambient Intell Human Comput

View full text Add to dashboard Cite

This article proposes a federated learning framework to build Random Forest, Support Vector Machine, and Linear Regression models for stock market prediction. The performance of the federated learning is compared against centralised and decentralised learning frameworks to figure out the best fitting approach for stock market prediction. According to the results, federated learning outperforms both centralised and decentralised frameworks in terms of Mean Square Error if Random Forest (MSE = 0.021) and Support Vector Machine techniques (MSE = 37.596) are used, while centralised learning (MSE = 0.011) outperforms federated and decentralised frameworks if a linear regression model is used. Moreover, federated learning gives a better model training delay as compared to the benchmarks if Linear Regression (time = 9.7 s) and Random Forest models (time = 515 s) are used, whereas decentralised learning gives a minimised model training delay (time = 3847 s) for Support Vector Machine.

show abstract

Machine learning concept in de-spiking process for nuclear resonant vibrational spectra – Automation using no external parameter

Cited by 6 publications

References 32 publications

Artificial Intelligence for Surface‐Enhanced Raman Spectroscopy

Artificial Intelligence for Surface‐Enhanced Raman Spectroscopy

X-ray and Nuclear Spectroscopies to Reveal the Element-Specific Oxidation States and Electronic Spin States for Nanoparticulated Manganese Cyanidoferrates and Analogs

A federated learning-enabled predictive analysis to forecast stock market trends

Contact Info

Product

Resources

About