Power Spectrum Analysis Research Articles

Examination of Magnetic Field Signatures and Local Plasma Distribution Variations in Jupiter's Magnetosphere

AbstractSpinning disk of plasma results in deformation of Jupiter's magnetic field from a dipole to a magnetodisc configuration with significant radial and azimuthal components. Polar orbit of the Juno spacecraft enables a comprehensive in situ observation of Jupiter's plasmadisc at different latitudes. Observations from magnetometer and plasma distribution instruments on Juno mission were used to examine interaction of magnetic field and heavy ions in Jupiter's nightside magnetosphere. Power spectrum analysis was used to evaluate magnetic fluctuations inside and outside of the plasmadisc. Extensive radial maps of magnetic activity in Jupiter's nightside magnetosphere are presented. Comprehensive examination of magnetic fluctuations enable to identify local plasma density variations that correlate with plasma distribution count rate fluctuations. Observations of magnetic signatures together with examination of plasma distribution are used to detect features that point to plasma transport inside and outside of Jupiter's plasmadisc. This work shows that plasma in Jupiter's magnetosphere is contained near centrifugal equator and is not a subject to curvature drift that could generate ring currents. Signatures described as current sheet crossings were investigated. On close inspection these signatures show attributes that belong to stand alone formations rather than a single current sheet structure. This work shows that current flow is not an adequate mechanism to explain deformation of Jupiter's magnetic field.

On the Detection and Characterization of Quasiperiodic Oscillations in Astronomical Time Series: Gamma-Ray Burst X-Ray Light Curves as a Test Case

The study of temporal properties of variable sources can elucidate their physical processes. In this context, we present a critical study comparing three approaches to periodic or quasiperiodic behavior: Gaussian process, power spectrum, and wavelet analysis, using celerite, Lomb–Scargle periodograms, and weighted wavelet Z-transforms, respectively. We use 15 Swift X-ray Telescope light curves of short gamma-ray bursts (sGRBs) as examples. A comprehensive analysis of two sGRB X-ray light curves is performed. The results reveal the importance of artifacts, largely in the form of false quasiperiodic oscillation signals, possibly introduced by preprocessing (such as detrending) or other aspects of the analysis. The exploration described in this paper can be helpful for future studies of variability in gamma-ray bursts, active galactic nuclei, and other astronomical sources.

The X-ray variability of active galactic nuclei: Power spectrum and variance analysis of the Swift/BAT light curves

Aims. We study the X-ray power spectrum of active galactic nuclei (AGN) in order to investigate whether Seyfert I and II power spectra are similar or not and whether AGN variability depends on the mass and accretion rate of black holes as well as to compare the power spectra of AGN with the power spectra of Galactic X-ray black hole binaries. Method. We used 14–195 keV band light curves from the 157-month Swift/BAT hard X-ray survey, and we computed the mean power spectrum and excess variance of AGN in narrow black hole mass and AGN luminosity bins. We fitted a power-law model to the AGN power spectra, and we investigated whether the power spectrum parameters and the excess variance depend on the black hole mass, luminosity, and accretion rate of AGN. Results. We found the Seyfert I and Seyfert II power spectra to be identical, in agreement with AGN unification models. The mean AGN X-ray power spectrum has the same power-law like shape, with a slope of −1 in all AGN irrespective of their luminosity and black hole mass. We did not detect any flattening to a slope of zero at frequencies as low as 10−9 Hz. We detected an anti-correlation between the power spectral density function (PSD) amplitude and the accretion rate, similar to what has been seen in the past in the 2–10 keV band. This implies that the variability amplitude in AGN decreases with an increasing accretion rate. The universal AGN power spectrum is consistent with the mean 2–9 keV band Cyg X-1 power spectrum in its soft state. We detected a small difference in amplitude, but this is probably due to the difference in energy. Conclusions. The mean low-frequency AGN X-ray power spectrum is consistent with the extension of the mean 0.01–25 Hz Cyg X-1 power spectrum in its soft state to lower frequencies. We cannot prove that the mean AGN PSD is analogous to the mean Cyg X-1 PSD in its soft state, as we do not know the location of the high-frequency break in the hard X-ray AGN PSDs. However, if this is the case, then the accretion disc in AGN probably extends to the radius of the innermost circular stable orbit (as is probably the case with the black hole binaries in their soft state). The X-ray corona will then be located on top, illuminating the disc and producing the X-ray reflection and disc reverberation phenomena commonly observed in these objects. Furthermore, the agreement between the PSD amplitude in AGN and the Cyg X-1 (either in the soft or the hard state) over many decades in frequency indicates that the X-ray variability process is probably the same in all accreting objects, irrespective of the mass of the compact object. We plan to investigate this issue further in the near future.

The Atacama Cosmology Telescope: Mitigating the Impact of Extragalactic Foregrounds for the DR6 Cosmic Microwave Background Lensing Analysis

We investigate the impact and mitigation of extragalactic foregrounds for the cosmic microwave background (CMB) lensing power spectrum analysis of Atacama Cosmology Telescope (ACT) data release 6 (DR6) data. Two independent microwave sky simulations are used to test a range of mitigation strategies. We demonstrate that finding and then subtracting point sources, finding and then subtracting models of clusters, and using a profile bias-hardened lensing estimator together reduce the fractional biases to well below statistical uncertainties, with the inferred lensing amplitude, A lens, biased by less than 0.2σ. We also show that another method where a model for the cosmic infrared background (CIB) contribution is deprojected and high-frequency data from Planck is included has similar performance. Other frequency-cleaned options do not perform as well, either incurring a large noise cost or resulting in biased recovery of the lensing spectrum. In addition to these simulation-based tests, we also present null tests on the ACT DR6 data for sensitivity of our lensing spectrum estimation to differences in foreground levels between the two ACT frequencies used, while nulling the CMB lensing signal. These tests pass whether the nulling is performed at the map or bandpower level. The CIB-deprojected measurement performed on the DR6 data is consistent with our baseline measurement, implying that contamination from the CIB is unlikely to significantly bias the DR6 lensing spectrum. This collection of tests gives confidence that the ACT DR6 lensing measurements and cosmological constraints presented in companion papers to this work are robust to extragalactic foregrounds.

Comparative electroencephalography analysis: Marathon runners during tapering versus sedentary controls reveals no significant differences.

Previous studies described various adaptive neuroplastic brain changes associated with physical activity (PA). EEG studies focused mostly on effects during or shortly after short bouts of exercise. This is the first study to investigate the capability of EEG to display PA-induced long-lasting plasticity in runners compared to a sedentary control group. Thirty trained runners and 30 age- and sex-matched sedentary controls (SC) were included as a subpopulation of the ReCaP (Running effects on Cognition and Plasticity) study. PA was measured with the International Physical Activity Questionnaire (IPAQ). Resting-state EEG of the runners was recorded in the tapering phase of the training for the Munich marathon 2017. Power spectrum analyses were conducted using standardized low-resolution electromagnetic tomography (sLORETA) and included the following frequency bands: delta: 1.5-6Hz, theta: 6.5-8.0Hz, alpha1: 8.5-10Hz, alpha2: 10.5-12.0Hz, beta1: 12.5-18.0Hz, beta2: 18.5-21.0Hz, beta3: 21.5-30.0Hz, and total power (1.5-30Hz). PA (IPAQ) and BMI differed significantly between the groups. The other included demographic parameters were comparable. Statistical nonparametric mapping showed no significant power differences in EEG between the groups. Heterogeneity in study protocols, especially in time intervals between exercise cessation and EEG recordings and juxtaposition of acute exercise-induced effects on EEG in previous studies, could be possible reasons for the differences in results. Future studies should record EEG at different time points after exercise cessation and in a broader spectrum of exercise intensities and forms to further explore the capability of EEG in displaying long-term exercise-induced plasticity.

Impact of astrophysical effects on the dark matter mass constraint with 21 cm intensity mapping

ABSTRACT We present an innovative approach to constraining the non-cold dark matter model using a convolutional neural network (CNN). We perform a suite of hydrodynamic simulations with varying dark matter particle masses and generate mock 21 cm radio intensity maps to trace the dark matter distribution at z = 3 in the postreionization epoch. Our proposed method complements the traditional power-spectrum analysis. We compare the results of the CNN classification between the mock maps with different dark matter masses with those from the two-dimensional power spectrum of the differential brightness temperature map of 21 cm radiation. We find that the CNN outperforms the power spectrum. Moreover, we investigate the impact of baryonic physics on the dark matter model constraint, including star formation, self-shielding of H i gas, and ultraviolet background model. We find that these effects may introduce some contamination in the dark matter constraint, but they are insignificant compared to the system noise of the SKA instruments.

Intraseasonal variability of the deep scattering layer induced by mesoscale eddy

The deep scattering layer (DSL), a stratum of the marine diel vertical migration (DVM) organisms inhabiting the mesopelagic ocean, plays a crucial role in transporting carbon and nutrients from the surface to depth through the migration of its organisms. Using 18 months of in-situ observations and altimeter sea level data, we reveal for the first time the intraseasonal variations and underlying mechanisms of the DSL and the DVM to the east of the Taiwan Island. Substantial vertical speeds acquired from the Acoustic Doppler Current Profiler were used to examine the distribution and variation of the DVM. Innovatively, the results for the power spectrum analysis of the scattering intensity demonstrated a significant intraseasonal variability (ISV) with an 80-day period in the DSL. Furthermore, the variation in the DVM was closely linked to the DSL and showed an 80-day ISV during the observation. A dynamic relationship between the ISV of the DSL east of Taiwan Island and the westward-propagating mesoscale eddies was established. Anticyclonic (cyclonic) eddy movement toward Taiwan Island triggers downward (upward) bending of the local isotherms, resulting in a layer of DSL warming (cooling) and subsequent upper boundary layer deepening (rising). These findings underscore the substantial influence of mesoscale eddies on biological activity in the mesopelagic ocean, establishing a novel understanding of ISV dynamics in the DSL and their links to eddy-induced processes.

Validation of spherical Fourier-Bessel power spectrum analysis with log-normal simulations and eBOSS DR16 LRG EZmocks

Validation of spherical Fourier-Bessel power spectrum analysis with log-normal simulations and eBOSS DR16 LRG EZmocks

Bayesian deep learning for cosmic volumes with modified gravity

Context. The new generation of galaxy surveys will provide unprecedented data that will allow us to test gravity deviations at cosmological scales at a much higher precision than could be achieved previously. A robust cosmological analysis of the large-scale structure demands exploiting the nonlinear information encoded in the cosmic web. Machine-learning techniques provide these tools, but no a priori assessment of the uncertainties. Aims. We extract cosmological parameters from modified gravity (MG) simulations through deep neural networks that include uncertainty estimations. Methods. We implemented Bayesian neural networks (BNNs) with an enriched approximate posterior distribution considering two cases: the first case with a single Bayesian last layer (BLL), and the other case with Bayesian layers at all levels (FullB). We trained both BNNs with real-space density fields and power spectra from a suite of 2000 dark matter-only particle-mesh N-body simulations including MG models relying on MG-PICOLA, covering 256 h−1 Mpc side cubical volumes with 1283 particles. Results. BNNs excel in accurately predicting parameters for Ωm and σ8 and their respective correlation with the MG parameter. Furthermore, we find that BNNs yield well-calibrated uncertainty estimates that overcome the over- and under-estimation issues in traditional neural networks. The MG parameter leads to a significant degeneracy, and σ8 might be one possible explanation of the poor MG predictions. Ignoring MG, we obtain a deviation of the relative errors in Ωm and σ8 by 30% at least. Moreover, we report consistent results from the density field and power spectrum analysis and comparable results between BLL and FullB experiments. This halved the computing time. This work contributes to preparing the path for extracting cosmological parameters from complete small cosmic volumes towards the highly nonlinear regime.

Aeroacoustic Loading of Impinging Supersonic Boundary-Layer Interaction on Statically Deformed Surfaces

This paper concerns the interaction of an impinging shock wave with a supersonic turbulent boundary layer over several distinct and permanently deformed surfaces, resulting in differences in the shock–boundary-layer interaction and the surface acoustic loading. High-order numerical simulations featuring two-dimensional surface deformations typically encountered in experiments are performed. The deformation amplitudes are up to half the incoming turbulent boundary-layer thickness. The results show that the high-pressure region about the shock impingement is significantly altered and can become narrower or wider depending on the local surface inclination of the deformed panel mode. The surface curvature is found to not significantly affect the separation and reattachment locations of the recirculation bubble. The power spectrum analysis of the pressure fluctuations along the panel’s midspan, where the surface attains the largest deformation amplitude, exhibits a rich and varied response. The pressure power spectrum is amplified in all of the surface deformation modes examined, with the magnitude of the amplification varying in the frequency domain, depending on the location and mode.

Effects of the Northern Hemisphere sudden stratospheric warmings on the Sporadic-E layers in the Brazilian sector

Tidal and Planetary Wave (PWs) amplitudes are strongly influenced by Sudden Stratospheric Warming (SSW) events. A nonlinear interaction between the tidal winds and planetary waves during the SSW may contribute to the intensification of sporadic-E (Es) layers in the lower thermosphere. This work investigated the relationship between SSW events in the Northern Hemisphere and the Es layer occurrence at low latitudes in the Brazilian sector. We used data from digital ionosondes installed in the observatories of Araguatins (ARA, 5.65° S; 48.12° W; dip lat. −5.44°) and São José dos Campos (SJC, 23.18° S; 45.89° W; dip lat. −21.37°) to analyze the Es layers. Additionally, we used the temperature, zonal wind, and PWs data at high latitudes in the Northern Hemisphere during the major SSW event that occurred in February/2018 and during the events of Dec/2018–Jan/2019 and Dec/2020–Jan/2021. The results showed a maximum frequency peak of 20 MHz (∼5 × 106 electrons.cm−3) at ARA and SJC during these SSW events. The large values of ftEs, fbEs, and electronic densities were observed between 100 and 115 km height in the Esf/l type layers during daytime or nighttime periods. The results also showed that the number of large values of ftEs, fbEs, and electronic density of the Es layer was much higher in ARA than in SJC, in general. The wavelet power spectrum analyses of the ftEs and fbEs showed a periodicity of 2-days before and after the central day of the SSWs events at the station of ARA, with three prominent peaks in the 2018/2019 event. At the SJC station the quasi-2-day periodicity in the wavelet analyses of the ftEs was observed after the central day in all three SSW events, with a peak before the central day during the 2020/2021 event.

Detection of long-range correlation causing multifractality in H time series of geomagnetic field over the Northern Hemisphere during quiet geomagnetic conditions

The generation of geomagnetic time series frequently involves intricate spatio-temporal dynamics, wherein nonlinearity and scaling emerge as pivotal mechanisms. The present work investigates the fractal properties of geomagnetic field, and its sources during quiet geomagnetic conditions, by analyzing the horizontal component data of 1 min resolution (H time series) from different latitude regions. Analysis using Auto Correlation Function (ACF), Power spectrum analysis (PSA), Rescaled range analysis (R/S) and Detrended Fluctuation Analysis (DFA) gave indications of the presence of long-range correlations, fractal properties and scaling behaviour of quiet period H time series data. The strong q dependence of generalized Hurst exponent h(q) and scaling exponent Ʈ(q), obtained from the Multifractal Detrended Fluctuation Analysis (MFDFA) method, confirmed the multifractal behaviour of quiet period H time series. The multifractal source analysis of the quiet period H time series with its shuffled and surrogate time series data revealed that, the multifractal characteristics are caused almost entirely by the long-range correlations between small and large fluctuations. The analysis of the latitudinal variations in the multifractal characteristics is also performed as a part of the work, by considering H time series belonging to various latitudes under almost same longitude. The observed decrease in the generalized Hurst exponent values with increase in latitude indicates the change in the multifractal characteristics of the time series with latitude. Also the work revealed that, the H time series from high latitude region shows anti-persistent behaviour, even when the H time series from observatories located at the low and mid latitude regions exhibiting persistent behaviour. The change in the persistence, anti-persistence behaviour can be related to the strength of Earth’s magnetic field.

High-skilled first-person shooting game players have specific frontal lobe activity: Power spectrum analysis in an electroencephalogram study

First-person shooting (FPS) games are among the most famous video games worldwide. However, cortical activities in environments related to real FPS games have not been studied. This study aimed to determine differences in cortical activity between low- and high-skilled FPS game players using 160-channel electroencephalography. Nine high-skilled FPS game players (official ranks: above the top 10%) and eight low-skilled FPS game players (official ranks: lower than the top 20%) were recruited for the experiment. The task was set for five different conditions using the AimLab program, which was used for the FPS game players’ training. Additionally, we recorded the brain activity in the resting condition before and after the task, in which the participants closed their eyes and relaxed. The reaction time and accuracy (the number of hit-and-miss targets) were calculated to evaluate the task performance. The results showed that high-skilled FPS game players have fast reaction times and high accuracy during tasks. High-skilled FPS game players had higher cortical activity in the frontal cortex than low-skilled FPS game players during each task. In low-skilled players, cortical activity level and performance level were associated. These results suggest that high cortical activity levels were critical to achieving high performance in FPS games.

Lower oddball event-related EEG delta and theta responses in patients with dementia due to Parkinson's and Lewy body than Alzheimer's disease

Oddball task-related EEG delta and theta responses are associated with frontal executive functions, which are significantly impaired in patients with dementia due to Parkinson's disease (PDD) and Lewy bodies (DLB). The present study investigated the oddball task-related EEG delta and theta responses in patients with PDD, DLB, and Alzheimer's disease dementia (ADD). During visual and auditory oddball paradigms, EEG activity was recorded in 20 ADD, 17 DLB, 20 PDD, and 20 healthy (HC) older adults. Event-related EEG power spectrum and phase-locking analysis were performed at the delta (1–4 Hz) and theta (4–7 Hz) frequency bands for target and nontarget stimuli. Compared to the HC persons, dementia groups showed lower frontal and central delta and theta power and phase-locking associated with task performance and neuropsychological test scores. Notably, this effect was more significant in the PDD and DLB than in the ADD. In conclusion, oddball task-related frontal and central EEG delta and theta responses may reflect frontal supramodal executive dysfunctions in PDD and DLB patients.

Enhanced adaptive high-frequency resonance technology for bearing fault detection

The traditional high-frequency resonance technology has two disadvantages: (1) The resonance frequency (RF) is obtained by using Fourier spectrum (FS) and the bandwidth is obtained based on experience, and the effect is often poor due to severe noise interference; (2) the separation of compound faults is not considered. After systematic research, a feasible fault detection methodology named enhanced adaptive high-frequency resonance technology is summarized and introduced, which is on the basis of power spectrum (PS) analysis, abbreviated as enhanced adaptive high-frequency resonance technique based on power spectrum (PS-EAHFRT), to realize the single fault detection and compound fault separation of bearing. Firstly, the theoretical values of various bearing fault modes are calculated to establish a database. Then, a series of spectral peaks are obtained by performing PS analysis, and each prominent spectral peak is considered as an alternative RF, that is, an alternative central frequency. Then, the bandwidth calculation formula of the bandpass filter is given, and the optimal bandwidth parameter is calculated by using the signal envelope kurtosis as a sparse measure. Finally, the original signal is denoised by the optimized band-pass filter, and then the filtered signal is demodulated by envelope PS analysis, and compared with the theoretical value in the library, and then the fault category is obtained. Simulated and real-world bearing cases are used to demonstrate the effectiveness and advantages of PS-EAHFRT. It is demonstrated that the PS-EAHFRT can effectively realize the bearing single fault detection and the compound fault separation. Moreover, the lower bound of the frequency interval in which this method can achieve bearing composite fault separation is explored.

The GLM-spectrum: A multilevel framework for spectrum analysis with covariate and confound modelling

Abstract The frequency spectrum is a central method for representing the dynamics within electrophysiological data. Some widely used spectrum estimators make use of averaging across time segments to reduce noise in the final spectrum. The core of this approach has not changed substantially since the 1960s, though many advances in the field of regression modelling and statistics have been made during this time. Here, we propose a new approach, the General Linear Model (GLM) Spectrum, which reframes time averaged spectral estimation as multiple regression. This brings several benefits, including the ability to do confound modelling, hierarchical modelling, and significance testing via non-parametric statistics. We apply the approach to a dataset of EEG recordings of participants who alternate between eyes-open and eyes-closed resting state. The GLM-Spectrum can model both conditions, quantify their differences, and perform denoising through confound regression in a single step. This application is scaled up from a single channel to a whole head recording and, finally, applied to quantify age differences across a large group-level dataset. We show that the GLM-Spectrum lends itself to rigorous modelling of within- and between-subject contrasts as well as their interactions, and that the use of model-projected spectra provides an intuitive visualisation. The GLM-Spectrum is a flexible framework for robust multilevel analysis of power spectra, with adaptive covariate and confound modelling.

Processing and re-interpretation of gravity bouguer map of a selected area in the W-NW of Iraq

Geophysical data interpretation is crucial in characterizing the subsurface structure. The Bouguer gravity map analysis of the W-NW region of Iraq serves as the basis for the current geophysical research. The Bouguer gravity data were processed using the Power Spectrum Analysis method. Four depth slices have been acquired after the PSA process, which are: 390 m, 1300 m, 3040 m, and 12600 m depth. The gravity anomaly depth maps show that shallow-depth anomalies are mainly related to the sedimentary cover layers and structures, while the gravity anomaly of the deeper depth slice of 12600 m is more presented to the basement rocks and mantle uplift. The 2D modeling technique was used for the quantitative interpretation of a selected Gravity profile along the study area. The model section of the gravity profile illustrated the relatively high density of subsurface basement rocks and/or upward mantle process which causes the effect of positive gravity values. Furthermore, several faults are indicated in the sedimentary sections by potential gravity methods such as several grabens, half grabens, and horst structures which are identified in Bouguer depth maps by relatively high and low gravity values, these structures were also affected by the basement rock uplift and/or Mantle upwards.

A Wind Field Reconstruction from Numerical Weather Prediction Data Based on a Meteo Particle Model

In the present work, a methodology for wind field reconstruction based on the Meteo Particle model (MPM) from numerical weather prediction (NWP) data is presented. The development of specific wind forecast services is a challenging research topic, in particular for what concerns the availability of accurate local weather forecasts in highly populated areas. Currently, even if NWP limited area models (LAMs) are run at a spatial resolution of about 1 km, this level of information is not sufficient for many applications; for example, to support drone operation in urban contexts. The coupling of the MPM with the NWP limited area model COSMO has been implemented in such a way that the MPM reads the NWP output over a selected area and provides wind values for the generic point considered for the investigation. The numerical results obtained reveal the good behavior of the method in reproducing the general trend of the wind speed, as also confirmed by the power spectra analysis. The MPM is able to step over the intrinsic limitations of the NWP model in terms of the spatial and temporal resolution, even if the MPM inherits the bias that inevitably affects the COSMO output.

Panic at the ISCO: Time-varying Double-peaked Broad Lines from Evolving Accretion Disks Are Common among Optically Variable AGNs

About 3%–10% of Type I active galactic nuclei (AGNs) have double-peaked broad Balmer lines in their optical spectra originating from the motion of gas in their accretion disk. Double-peaked profiles arise not only in AGNs, but occasionally appear during optical flares from tidal disruption events and changing-state AGNs. In this paper, we identify 250 double-peaked emitters (DPEs) among a parent sample of optically variable broad-line AGNs in the Zwicky Transient Facility (ZTF) survey, corresponding to a DPE fraction of 19%. We model spectra of the broad Hα emission-line regions and provide a catalog of the fitted accretion disk properties for the 250 DPEs. Analysis of power spectra derived from the 5 yr ZTF light curves finds that DPE light curves have similar amplitudes and power-law indices to other broad-line AGNs. Follow-up spectroscopy of 12 DPEs reveals that ∼50% display significant changes in the relative strengths of their red and blue peaks over long 10–20 yr timescales, indicating that broad-line profile changes arising from spiral arm or hotspot rotation are common among optically variable DPEs. Analysis of the accretion disk parameters derived from spectroscopic modeling provides evidence that DPEs are not in a special accretion state, but are simply normal broad-line AGNs viewed under the right conditions for the accretion disk to be easily visible. We include inspiraling supermassive black hole binary candidate SDSSJ1430+2303 in our analysis, and discuss how its photometric and spectroscopic variability is consistent with the disk-emitting AGN population in the ZTF survey.

A non-linear grey Fourier model based on kernel method for seasonal traffic speed forecasting

This article introduces a non-linear grey Fourier model utilizing the kernel method to better characterize seasonal traffic speed with fluctuating trends. The linear trend term in the grey Fourier model is replaced with feature mapping, resulting in a non-linear form. Utilizing Lagrange multipliers and a kernel function, the parameter estimation is transformed into a linear equation-solving problem. Order selection is carried out via power spectrum analysis, and hyper-parameters are determined using Bayesian optimization. Two numerical examples from real problems are presented to validate the model’s performance. The results indicate that the developed model outperforms nine models across four categories. Specifically, the model is applied to predict traffic speed in Guangzhou, and comparative results with four neural networks demonstrate its commendable performance.