Search

Tweek atmospherics are extremely low frequency and very low frequency pulse signals with frequency dispersion characteristics that originate from lightning discharges and that propagate in the Earth–ionosphere waveguide over long distances. In this study, we developed an automatic method to recognize tweek atmospherics and diagnose the lower ionosphere based on the machine learning method. The differences (automatic − manual) in each ionosphere parameter between the automatic method and the manual method were −0.07 ± 2.73 km, 0.03 ± 0.92 cm−3, and 91 ± 1,068 km for the ionospheric reflection height (h), equivalent electron densities at reflection heights (Ne), and propagation distance (d), respectively. Moreover, the automatic method is capable of recognizing higher harmonic tweek sferics. The evaluation results of the model suggest that the automatic method is a powerful tool for investigating the long-term variations in the lower ionosphere.

With the development of globalization, people’s communication becomes more and more frequent. China urgently needs to cultivate talents at the international level and strong foreign language communicative ability to participate in international affairs and follow the international development trend. Based on the data analysis method and machine learning method in big data analysis, this article explores the current situation of intercultural communication ability and competence cultivation of language students in colleges and universities. The article summarizes the existing problems of language students’ intercultural communication ability in Chinese universities based on questionnaires conducted in four universities in southwest, northeast, east, and South China. Secondly, based on the current problems in the intercultural communication competence of language students in Chinese universities, the article further proposes teaching inspirations for the cultivation of intercultural communication competence of language students in Chinese universities. The survey results show that the highest total number of respondents who chose A and B answers, i.e., they think they have strong intercultural communication ability, was 689 from Zhongshan University, accounting for 33.62%, followed by 615 from Shanghai University, accounting for 30.1%; the lowest was 240 from Guizhou University, accounting for 11.7% only, and the level of intercultural communication ability of language students in these four universities can be ranked as follows. Zhongshan University > Shanghai University > Heilongjiang University > Guizhou University. This study can suggest highly rigorous and practical strategies for cultivating language students’ intercultural communication ability in Chinese universities.

Background: Aid volatility has been frequently discussed but its effect on health poverty has not. Aim: The aim of this study is to evaluate the effect of aid volatility on residential health in recipient countries. Setting: Panel data from 91 recipient countries around the world from 1989 to 2017 were analysed. Method: The Random Forest algorithm in machine learning was used. Results: It reveals a conductive effect of aid volatility on public health, as aid volatility accounts for approximately 2.2%, 3.6%, 2.5% significance in child mortality, maternal mortality and life expectancy, respectively. Positive volatility causes more child mortality than negative volatility, while it causes less maternal mortality, although both kinds of volatility have a significant effect on life expectancy. Additionally, multinational aid volatility has a greater impact than bilateral volatility. The age dependency ratio affects maternal health and public life expectancy, while foreign aid and aid volatility are not crucial. Conclusion: The donor nations or institutions should keep their aid growth-rate stable in order to ensure sustainable development in both the local economy and public health, while the recipient nations are encouraged to address these issues by developing their economy internally. Contribution: According to a conductive effect framework developed during the research, theoretically, an indirect relationship between aid volatility and the health poverty level of recipient countries exists.

Background: In patients with cardiac arrest who remain comatose after return of spontaneous circulation, seizures and other abnormalities on electroencephalogram (EEG) are common. Thus, guidelines recommend urgent initiation of EEG for the evaluation of seizures in this population. Point-of-care EEG systems, such as Ceribell™ Rapid Response EEG (Rapid-EEG), allow for prompt initiation of EEG monitoring, albeit through a reduced-channel montage. Rapid-EEG incorporates an automated seizure detection software (Clarity™) to measure seizure burden in real time and alert clinicians at the bedside when a high seizure burden, consistent with possible status epilepticus, is identified. External validation of Clarity is still needed. Our goal was to evaluate the real-world performance of Clarity for the detection of seizures and status epilepticus in a sample of patients with cardiac arrest. Methods: This study was a retrospective review of Rapid-EEG recordings from all the patients who were admitted to the medical intensive care unit at Kent Hospital (Warwick, RI) between 6/1/2021 and 3/18/2022 for management after cardiac arrest and who underwent Rapid-EEG monitoring as part of their routine clinical care (n = 21). Board-certified epileptologists identified events that met criteria for seizures or status epilepticus, as per the 2021 American Clinical Neurophysiology Society's Standardized Critical Care EEG Terminology, and evaluated any seizure burden detections generated by Clarity. Results: In this study, 4 of 21 patients with cardiac arrest (19.0%) who underwent Rapid-EEG monitoring had multiple electrographic seizures, and 2 of those patients (9.5%) had electrographic status epilepticus within the first 24 h of the study. None of these ictal abnormalities were detected by the Clarity seizure detection system. Clarity showed 0% seizure burden throughout the entirety of all four Rapid-EEG recordings, including the EEG pages that showed definite seizures or status epilepticus. Conclusions: The presence of frequent electrographic seizures and/or status epilepticus can go undetected by Clarity. Timely and careful review of all raw Rapid-EEG recordings by a qualified human EEG reader is necessary to guide clinical care, regardless of Clarity seizure burden measurements. [ABSTRACT FROM AUTHOR]

The leaf area index (LAI) is an essential biophysical parameter for describing the vegetation canopy structure and predicting its growth and productivity. Using unmanned aerial vehicle (UAV) hyperspectral imagery to accurately estimate the LAI is of great significance for Spartina alterniflora (S. alterniflora) growth status monitoring. In this study, UAV hyperspectral imagery and the LAI of S. alterniflora during the flourishing growth period were acquired. The hyperspectral data were preprocessed with Savitzky–Golay (SG) smoothing, and the first derivative (FD) and the second derivative (SD) spectral transformations of the data were then carried out. Then, using the band combination index (BCI) method, the characteristic bands related to the LAI were extracted from the hyperspectral image data obtained with the UAV, and spectral indices (SIs) were constructed through the characteristic bands. Finally, three machine learning (ML) regression methods—optimized support vector regression (OSVR), optimized random forest regression (ORFR), and optimized extreme gradient boosting regression (OXGBoostR)—were used to establish LAI estimation models. The results showed the following: (1) the three ML methods accurately predicted the LAI, and the optimal model was provided by the ORFR method, with an R2 of 0.85, an RMSE of 0.19, and an RPD of 4.33; (2) the combination of FD SIs improved the model accuracy, with the R2 value improving by 41.7%; (3) the band combinations screened using the BCI method were mainly concentrated in the red and near-infrared bands; (4) the higher LAI was distributed on the seaward side of the study area, while the lower LAI was located at the junction between the S. alterniflora and the tidal flat. This study serves as both theoretical and technological support for research on the LAI of S. alterniflora and as a solid foundation for the use of UAV remote sensing technologies in the supervisory control of S. alterniflora. [ABSTRACT FROM AUTHOR]

River transport of dissolved organic carbon (DOC) to the ocean is a crucial but poorly quantified regional carbon cycle component. Large uncertainties remaining on the riverine DOC export from China, as well as its trend and drivers of change, have challenged the reconciliation between atmosphere‐based and land‐based estimates of China's land carbon sink. Here, we harmonized a large database of riverine in‐situ measurements and applied a random forest model, to quantify riverine DOC fluxes (FDOC) and DOC concentrations (CDOC) in rivers across China. This study proposes the first DOC modeling effort capable of reproducing well the magnitude of riverine CDOC and FDOC, as well as its trends, on a monthly scale and with a much wider spatial distribution over China compared to previous studies that mainly focused on annual‐scale estimates and large rivers. Results show that over the period 2001–2015, the average CDOC was 2.25 ± 0.45 mg/L and average FDOC was 4.04 ± 1.02 Tg/year. Simultaneously, we found a significant increase in FDOC (+0.044 Tg/year2, p =.01), but little change in CDOC (−0.001 mg/L/year, p >.10). Although the trend in CDOC is not significant at the country scale, it is significantly increasing in the Yangtze River Basin and Huaihe River Basin (0.005 and 0.013 mg/L/year, p <.05) while significantly decreasing in the Yellow River Basin and Southwest Rivers Basin (−0.043 and −0.014 mg/L/year, p =.01). Changes in hydrology, play a stronger role than direct impacts of anthropogenic activities in determining the spatio‐temporal variability of FDOC and CDOC across China. However, and in contrast with other basins, the significant increase in CDOC in the Yangtze River Basin and Huaihe River Basin is attributable to direct anthropogenic activities. Given the dominance of hydrology in driving FDOC, the increase in FDOC is likely to continue under the projected increase in river discharge over China resulting from a future wetter climate. [ABSTRACT FROM AUTHOR]

With the rapid development of the economy, it is important to reasonably evaluate the development status of the regional manufacturing industry. Given this, this article expands the evaluation indicators of urban advanced manufacturing (UAM) from the perspective of the push–pull-mooring (PPM). Then, it uses a machine learning (ML) method to predict the evaluation results of other cities through a small amount of sample data. The results show that: (1) From the current development status of UAM in Guangdong Province (GD), cities in the Pearl River Delta region occupy a dominant position. However, cities in eastern, western, and mountainous regions have strong development potential and lead cities. Therefore, each region has cities with high levels of development and has a demonstrative role. (2) By comparison, it was found that the overall development level of UAM in GD is not significantly different from that of the Yangtze River Economic Belt. However, due to significant differences in their extreme values, the proportion of cities above the average in the overall population is relatively small. This indirectly proves that GD's UAM not only has a phased nature, but also has a demonstrative role. (3) The prediction effect of the perceptron model is better than other methods. Although neural network models have better prediction performance than other machine learning models, they should not overly rely on complex network structure prediction data. By comparing the results, the reliability is verified. Finally, according to the life cycle theory, we propose a targeted development path for different UAM. [ABSTRACT FROM AUTHOR]

Since fossil fuels are slowly depleting, bio and renewable energies are now given more attention. The main purpose of this research is to investigate and optimize the influencing parameters of bioenergy production through transesterification process. The application of artificial intelligence (AI) in bioenergy production studies has become increasingly popular due to its capability of interpreting nonlinear relationships between inputs and outputs for complex systems. Here, after conducting library studies and carefully reviewing the existing methods, the multi-layer perceptron (MLP), K-nearest neighbors (KNN), Artificial neural network (ANN), and Gaussian processes regression (GPR) models were selected for simulation and prediction of the efficiency of fatty acid methyl ester (FAME) production. The main effective transesterification parameters on production of biodiesel including the temperature of reaction (°C), catalyst mass to oil mass ratio (wt.%), and the molar ratio of methanol to oil were set as the input variables in all studied models. For reaction between oil and short chain alcohols, wollastonite (a calcium metasilicate, CaSiO3) was utilized as a phase boundary catalyst. By carefully selecting the execution conditions of the algorithms in the model selection phase, all three models reached a result above 0.99 and close to 1 with the square R criterion. Also, the RMSE values for the studied models were 3.95 for MLP, 1.09 for KNN, 0.13 for ANN and 3.60 for GPR models. Therefore, it can be concluded that although the ANN model was to be a better model in process efficiency prediction in terms of error, but all three algorithms had high accuracy because of different generality types. The optimum yield of 97.8% for FAME production was observed at optimum methanol to oil molar ratio, reaction temperature, and catalyst mass to oil mass ratio 65°C, 15, and 9.21 wt%, respectively.

The torque ripples in a switched reluctance motor (SRM) are minimized via an optimal adaptive dynamic regulator that is presented in this research. A novel reinforcement neural network learning approach based on machine learning is adopted to find the best solution for the tracking problem of the SRM drive in real time. The reference signal model which minimizes the torque pulsations is combined with tracking error to construct the augmented structure of the SRM drive. A discounted cost function for the augmented SRM model is described to assess the tracking performance of the signal. In order to track the optimal trajectory, a neural network (NN)-based RL approach has been developed. This method achieves the optimal tracking response to the Hamilton–Jacobi–Bellman (HJB) equation for a nonlinear tracking system. To do so, two neural networks (NNs) have been trained online individually to acquire the best control policy to allow tracking performance for the motor. Simulation findings have been undertaken for SRM to confirm the viability of the suggested control strategy. [ABSTRACT FROM AUTHOR]

Histone deacetylase (HDAC) 1, a member of the histone deacetylases family, plays a pivotal role in various tumors. In this study, we collected 7313 human HDAC1 inhibitors with bioactivities to form a dataset. Then, the dataset was divided into a training set and a test set using two splitting methods: (1) Kohonen's self-organizing map and (2) random splitting. The molecular structures were represented by MACCS fingerprints, RDKit fingerprints, topological torsions fingerprints and ECFP4 fingerprints. A total of 80 classification models were built by using five machine learning methods, including decision tree (DT), random forest, support vector machine, eXtreme Gradient Boosting and deep neural network. Model 15A_2 built by the XGBoost algorithm based on ECFP4 fingerprints showed the best performance, with an accuracy of 88.08% and an MCC value of 0.76 on the test set. Finally, we clustered the 7313 HDAC1 inhibitors into 31 subsets, and the substructural features in each subset were investigated. Moreover, using DT algorithm we analyzed the structure–activity relationship of HDAC1 inhibitors. It may conclude that some substructures have a significant effect on high activity, such as N-(2-amino-phenyl)-benzamide, benzimidazole, AR-42 analogues, hydroxamic acid with a middle chain alkyl and 4-aryl imidazole with a midchain of alkyl whose α carbon is chiral. [ABSTRACT FROM AUTHOR]

In the process of ocean exploration, highly accurate and sensitive measurements of seawater temperature and pressure significantly impact the study of seawater's physical, chemical, and biological processes. In this paper, three different package structures, V-shape, square-shape, and semicircle-shape, are designed and fabricated, and an optical microfiber coupler combined Sagnac loop (OMCSL) is encapsulated in these structures with polydimethylsiloxane (PDMS). Then, the temperature and pressure response characteristics of the OMCSL, under different package structures, are analyzed by simulation and experiment. The experimental results show that structural change hardly affects temperature sensitivity, and square-shape has the highest pressure sensitivity. In addition, with an input error of 1% F.S., temperature and pressure errors were calculated, which shows that a semicircle-shape structure can increase the angle between lines in the sensitivity matrix method (SMM), and reduce the effect of the input error, thus optimizing the ill-conditioned matrix. Finally, this paper shows that using the machine learning method (MLM) effectively improves demodulation accuracy. In conclusion, this paper proposes to optimize the ill-conditioned matrix problem in SMM demodulation by improving sensitivity with structural optimization, which essentially explains the cause of the large errors for multiparameter cross-sensitivity. In addition, this paper proposes to use the MLM to solve the problem of large errors in the SMM, which provides a new method to solve the problem of the ill-conditioned matrix in SMM demodulation. These have practical implications for engineering an all-optical sensor that can be used for detection in the ocean environment. [ABSTRACT FROM AUTHOR]

We present a retrospective analysis of Czech anti-covid governmental measures' effectiveness for an unusually long three years of observation. Numerous Czech government restrictive measures illustrate this analysis applied to three years of COVID-19 data from the first three COVID-19 cases detected on 1st March 2020 till March 2023. It illustrates the course from the dramatic combat of unknown illness to resignation to country-wide measures and placing COVID-19 into a category of common nuisances. Our analysis uses the derived adaptive recursive Bayesian stochastic multidimensional Covid model-based prediction of nine essential publicly available COVID-19 data series. The COVID-19 model enables us to differentiate between effective measures and solely nuisance or antagonistic provisions and their correct or wrong timing. Our COVID model allows us to predict vital covid statistics such as the number of hospitalized, deaths, or symptomatic individuals, which can serve for daily control of anti-covid measures and the necessary precautions and formulate recommendations to control future pandemics. [ABSTRACT FROM AUTHOR]

Complexity and unpredictability nature of earthquakes makes them unique external loads that there is no unique formula used for the prediction of seismic responses. Hence, this research aims to implement the most well-known Machine Learning (ML) methods in Python software to propose a prediction model for seismic response and performance assessment of Reinforced Concrete Moment-Resisting Frames (RC MRFs). To prepare 92,400 data points of training dataset for developing data-driven techniques, Incremental Dynamic Analyses (IDAs) were performed considering 165 RC MRFs with two-, to twelve-Story elevations having the bay lengths of 5.0 m, 6.1 m, and 7.6 m assuming near-fault seismic excitations. Then, important structural features were considered in datasets to train and test the ML-based prediction models, which were improved with innovative techniques. The results show that improved algorithms have higher R2 values for estimating the Maximum Interstory Drift Ratio (IDRmax), and two improved algorithms of artificial neural networks and extreme gradient boosting can estimate the Median of IDA curves (M-IDAs) of RC MRFs, which can be used to estimate the seismic limit-state capacity and performance assessment of existing or newly constructed RC buildings. To validate the generality and accuracy of the proposed ML-based prediction model, a five-Story RC building with different input features was used, and the results are promising. Therefore, graphical user interface is introduced as user-friendly tool to help researchers in estimating the seismic limit-state capacity of RC buildings, while reducing the computational cost and analytical efforts. [ABSTRACT FROM AUTHOR]

An emerging real-time ground compaction and quality control, known as intelligent compaction (IC), has been applied for efficiently optimising the full-area compaction. Although IC technology can provide real-time assessment of uniformity of the compacted area, accurate determination of the soil stiffness required for quality control and design remains challenging. In this paper, a novel and advanced numerical model simulating the interaction of vibratory drum and soil beneath is developed. The model is capable of evaluating the nonlinear behaviour of underlying soil subjected to dynamic loading by capturing the variations of damping with the cyclic shear strains and degradation of soil modulus. The interaction of the drum and the soil is simulated via the finite element method to develop a comprehensive dataset capturing the dynamic responses of the drum and the soil. Indeed, more than a thousand three-dimensional (3D) numerical models covering various soil characteristics, roller weights, vibration amplitudes and frequencies were adopted. The developed dataset is then used to train the inverse solver using an innovative machine learning approach, i.e. the extended support vector regression, to simulate the stiffness of the compacted soil by adopting drum acceleration records. Furthermore, the impacts of the amplitude and frequency of the vibration on the level of underlying soil compaction are discussed. The proposed machine learning approach is promising for real-time extraction of actual soil stiffness during compaction. Results of the study can be employed by practising engineers to interpret roller drum acceleration data to estimate the level of compaction and ground stiffness during compaction.

Accurate monitoring of the spatiotemporal variations in the gross primary productivity (GPP) in coastal wetlands is essential for blue carbon quantification. However, the currently available moderate-spatial-resolution GPP algorithms and products may contain large uncertainties and may rarely meet the demands of tidal wetland research. In this paper, we examine two statistical methods, namely, simple linear regression and random forest regression, to capture the high spatial resolution GPP variations in the salt marshes in the Yangtze Estuary by integrating multi-source data, including eddy covariance, Landsat, and Moderate-resolution Imaging Spectroradiometer (MODIS) satellite data. The satellite-derived photosynthetically active radiation (PAR), enhanced vegetation index (EVI), normalized difference vegetation index (NDVI), and normalized difference water index (NDWI) are used individually and in different combinations to drive the two statistical methods and investigate their performances in estimating the GPP. The results show that together with the PAR, the EVI generally has the greatest potential for GPP estimation (R2 > 0.75, RMSE

There is significant enterohepatic circulation (EHC) during the disposition of mycophenolic acid (MPA). The aim of this study was to elucidate factors influencing the EHC of MPA in Chinese adult renal allograft recipients. After 2 weeks of therapy with mycophenolate mofetil or enteric‐coated mycophenolate sodium, blood samples were collected from 125 patients at 0 to 12 hours post‐administration and MPA concentrations were determined. The influence of calcineurin inhibitors (CNIs) and genetic polymorphisms on MPA exposure and EHC was studied. The Shapley additive explanations method was used to estimate the impact of various factors on the area under the plasma drug concentration–time curve (AUC0–12h) for MPA. An extreme gradient boosting (XGboost) machine learning‐based model was established to predict AUC0–12h. Results showed that the dose‐normalized AUC6–12h (dn‐AUC6–12h) of MPA was significantly lower in patients co‐administered with cyclosporine (CsA) than in patients co‐administered with tacrolimus (TAC) (P <.05). For patients co‐administered with TAC, patients with ABCC2 C‐24T CC or SLCO1B1 T521C TT genotypes had significantly higher values of dn‐AUC6‐12h (P <.05). Patients with SLCO1B3 334T/699G alleles had significantly lower dn‐AUC6–12h values than homozygotes (P <.05). By introducing body weight, age, and EHC‐related factors, including co‐administered CNIs and genetic polymorphism of drug transporters, as covariates in the XGboost machine learning model, the prediction performance of AUC0–12h for MPA in Chinese adult renal allograft recipients can be improved. [ABSTRACT FROM AUTHOR]

Multiple types of COVID-19 vaccines have been shown to be highly effective in preventing SARS-CoV-2 infection and in reducing post-infection symptoms. Almost all of these vaccines induce systemic immune responses, but differences in immune responses induced by different vaccination regimens are evident. This study aimed to reveal the differences in immune gene expression levels of different target cells under different vaccine strategies after SARS-CoV-2 infection in hamsters. A machine learning based process was designed to analyze single-cell transcriptomic data of different cell types from the blood, lung, and nasal mucosa of hamsters infected with SARS-CoV-2, including B and T cells from the blood and nasal cavity, macrophages from the lung and nasal cavity, alveolar epithelial and lung endothelial cells. The cohort was divided into five groups: non-vaccinated (control), 2*adenovirus (two doses of adenovirus vaccine), 2*attenuated (two doses of attenuated virus vaccine), 2*mRNA (two doses of mRNA vaccine), and mRNA/attenuated (primed by mRNA vaccine, boosted by attenuated vaccine). All genes were ranked using five signature ranking methods (LASSO, LightGBM, Monte Carlo feature selection, mRMR, and permutation feature importance). Some key genes that contributed to the analysis of immune changes, such as RPS23, DDX5, PFN1 in immune cells, and IRF9 and MX1 in tissue cells, were screened. Afterward, the five feature sorting lists were fed into the feature incremental selection framework, which contained two classification algorithms (decision tree [DT] and random forest [RF]), to construct optimal classifiers and generate quantitative rules. Results showed that random forest classifiers could provide relative higher performance than decision tree classifiers, whereas the DT classifiers provided quantitative rules that indicated special gene expression levels under different vaccine strategies. These findings may help us to develop better protective vaccination programs and new vaccines. [ABSTRACT FROM AUTHOR]

The structural complexities of grain boundaries (GBs) result in their complicated property contributions to polycrystalline metals and alloys. In this study, we propose a GB structure descriptor by linearly combining the average two-point correlation function (PCF) and standard deviation of PCF via a weight parameter, to reveal the standard deviation effect of PCF on energy predictions of Cu, Al and Ni asymmetric tilt GBs (i.e., Σ3, Σ5, Σ9, Σ11, Σ13 and Σ17), using two machine learning (ML) methods; i.e., principal component analysis (PCA)-based linear regression and recurrent neural networks (RNN). It is found that the proposed structure descriptor is capable of improving GB energy prediction for both ML methods. This suggests the discriminatory power of average PCF for different GBs is lifted since the proposed descriptor contains the data dispersion information. Meanwhile, we also show that GB atom selection methods by which PCF is evaluated also affect predictions. [ABSTRACT FROM AUTHOR]

• Applications of characterization approaches in asphalt performance prediction were reviewed. • The relationship between MD simulation parameters and macro-properties was summarized. • The potential links between micro–macro scales were discussed. • The combination of characterization and simulation approaches was prospected. Examining the intricate relationship between micro-components and macro-performance represents a pivotal focus within asphalt materials research. This review offers a comprehensive discussion on the micro–macro correlations in asphalt, delineating these correlations from two primary perspectives: characterization approaches and molecular dynamic (MD) simulation. This review begins by outlining foundational principles and assessment methodologies concerning saturates, aromatics, resins, and asphaltenes (SARA) components separation method, as well as techniques such as Fourier-transform infrared spectroscopy (FTIR), nuclear magnetic resonance (NMR), and gel permeation chromatography (GPC) methods. The application of the above methods in examining the relationship between chemical structure, components, and performance is highlighted, offering both qualitative and quantitative insights. The review then considers how MD simulation contributes to the investigation of SARA fractions interactions and analyses of physical, rheological, self-healing, and interfacial adhesion performance in asphalt materials, within the broader context of the micro–macro links. Finally, a proposition is forwarded to amalgamate characterization methodologies with data obtained from MD simulations, aiming to establish a comprehensive material information database. This initiative seeks to develop a robust structure–activity relationship model for asphalt, utilizing advanced machine learning (ML) methodologies. This review not only advances comprehension regarding the intricate components of asphalt but also furnishes a foundation for the targeted design and optimization of superior-quality asphalt materials. [ABSTRACT FROM AUTHOR]

• Foam-filled multi-layer lattice composite structures exhibited great energy absorption performance. • A novel framework for predicting the impact force-time series of composite structures was proposed. • Dynamic responses of proposed composite structures were examined in time and frequency domains. • Sensitivity analysis revealed that a high peak impact force corresponds to a short impact duration. Numerical simulations can provide valuable insights for the optimization of design and operational management; however, they are often impractical and computationally intensive. Machine learning methods are appealing to these problems due to their sufficient efficiency and accuracy. In this study, a novel framework for predicting the impact responses of foam-filled multi-layer lattice composite structures (FMLCSs) was proposed by combining the accurate finite element (FE) analyses, surrogate models, fast Fourier transform (FFT) method, and inverse FFT (IFFT) method. Firstly, reliable FM models were established to simulate the crashworthiness of the five FMLCSs under impact loading, including an analysis of energy transformation. Subsequently, surrogate models, namely radial basis function (RBF), polynomial response surface (PRS), Kriging (KRG), and back propagation neural network (BPNN), combined with methods of FFT and IFFT, were employed to predict the impact force-time series of the FMLCSs. More than 1000 frequency points were employed for each type of FMLCS, and all the R-square (R 2) values of the established surrogate models exceeded 0.95, indicating that the proposed framework accurately predicted the impact duration and impact responses in the frequency domain. In addition, parameter sensitivity analysis revealed that a high peak impact force was accompanied by a short impact duration. Moreover, increasing the lattice-web height resulted in a significant increase in the impact duration. [ABSTRACT FROM AUTHOR]

To rapidly realize the identification of fresh milk for water adulteration and to predict the amount of water adulteration, this study adopts a coaxial probe and a vector network analyzer to analyze the variation laws of dielectric constant ε', dielectric loss factor ε'' and dielectric loss angle tangent tanδ in the range of 2–20 GHz under 100 frequency points. Soft independent modeling of class analogy (SIMCA), Naive Bayes (NB), K-nearest neighbors (KNN), and support vector machine (SVM) models were built based on full dielectric spectra for qualitative testing of adulteration levels in milk. Feature variables of FDS were extracted by using the successive projections algorithm (SPA) and uninformative variables elimination (UVE). Partial least squares regression (PLSR), support vector regression (SVR), and particle swarm optimization least square support vector regression (PSO-LSSVR) models were built for quantitative testing of adulteration levels in milk. The results demonstrate an increasing trend of ε'' and tanδ with increasing frequency and a decreasing trend of ε'. The ε'-SIMCA model achieves the best effect in distinguishing water adulteration in milk, showing accuracy (ACC) of 1, Sensitivity (SNS) of 1, Specificity (SPC) precision (PRE) of 1, and an F 1 -score (F 1) of 1. The tanδ-SPA-PSO-LSSVR model optimally predicts the optimal prediction effect of moisture content of milk, showing an R P 2 of 0.994 and an RMSEP of 0.016 %. This study is conducive to building a more comprehensive knowledge system of water adulteration in fresh milk. Its results can provide a theoretical reference in the development of non-destructive detection instruments for natural milk. • Non-destructive detection of water adulteration in milk realized. • Models were built based on full dielectric spectra for qualitative testing of adulteration levels in milk. • Models were built to predict the level of milk adulteration based on extracted features from the full dielectric spectra. • Qualitative discriminative modeling accuracy up to 0.981, quantitative prediction modeling accuracy up to 0.994. [ABSTRACT FROM AUTHOR]

In this work, the vacuum arc-melting was used to prepare the Ti-10V-5Al-2.5Fe-0.1B alloy. Single-pass isothermal compression experiments were carried out on the alloy in the temperature range of 770–920 °C at strain rates of 0.0005–0.5 s−1. The BP model optimized by the bald eagle search algorithm (BES-BP), the BP model optimized by the sparrow search algorithm (SSA-BP), and the BP model optimized by the gray wolf optimization algorithm (GWO-BP) were developed for high-precision prediction of flow stress. The above models were compared by using the mean square correlation coefficient, root mean square error, and average absolute relative error between the predicted and experimental flow stress. The three prediction accuracy parameters have indicated that the BES-BP model has a higher accuracy for flow stress prediction at the known and the new process parameters. A hot processing map based on the dynamic materials model was developed by using the flow stress predicted in the framework of the BES-BP model, and EBSD analysis was performed as well. The results show that the degree of dynamic recrystallization increases with an increase in the power dissipation factor, and the formation of deformation bands is the main cause of instability. The minimum critical stress for inducing dynamic recrystallization of the alloy was found to be 13.13 MPa at 890 °C/0.0005 s−1. Moreover, the power dissipation factor increases with a decrease in critical stress. In addition, microstructure validation data reveal that the dynamic recrystallization model has a high accuracy for critical stress prediction, confirming that the critical stress increases with a decrease in the dynamic recrystallization fraction. • A new two-phase titanium alloy Ti-10V-5Al-2.5Fe-0.1B as an experimental material. • The bald eagle search algorithm, he sparrow search algorithm , and the gray wolf optimization algorithm were used to optimize the BP neural network. • The prediction accuracy of the BES-BP model is higher than that of the SSA-BP model and the GWO-BP model. • Different regions of microstructure evolution were verfied for the accuracy of the hot processing map. • Microstructural validation was carried out to verify the accuracy of the dynamic recrystallization critical model. [ABSTRACT FROM AUTHOR]

In the middle and late stages of the production of strong bottom water sandstone reservoirs, entering the medium and high water content period, the bottom water cone inlet phenomenon is prominent, the output is seriously decreasing, and it is necessary to take drainage control cone and other flow control field measures to maintain normal production, and clarify the main control influencing factors of the water content rise law of the reservoir and the water content rise mode can effectively guide the implementation of the process measures of the flow control field. Based on the production status of the nine areas of Tahe Oilfield, the water production of horizontal wells in the well area was statistically calculated, and the pattern of water rise of horizontal wells in the well area at this stage was summarized, and according to the geological and engineering characteristics of the well area, combined with the law mode of water content rise under different production stages, it was qualitatively analyzed in time and space. The main control factors of the rising law of reservoir water content were calculated in combination with the grey correlation method and machine learning method, and the correlation degree of each influencing factor in different production stages in different regions was calculated, and the main control factor of the rising law of oil reservoir water content was analyzed in time and space. The results show that at present, the patterns of water content rise of horizontal wells in the ninth area of Tahe oilfield are four types: climbing type, ladder type, rapid flooding type and open well flooding type. Combined with qualitative and quantitative analysis, the main controlling factors of the rising law of water content in the early stage of production: the structural position, the angle between the horizontal well section of the well and the direction of the ancient river channel in the water avoidance height, and the law of the rise of water content in the middle of the mining period is mainly spread by the interlayer within the layer. Reservoir heterogeneity is supplemented by late flow field regulation is mainly affected by reservoir heterogeneity and interlayer spread. It provides certain theoretical guidance for the implementation of process measures such as liquid extraction control and flow field regulation under different production stage conditions. [ABSTRACT FROM AUTHOR]

Drought is a major disaster over the Tibetan Plateau (TP) that exerts great impacts on natural ecosystems and agricultural production. Furthermore, most drought indices are only useful for assessing drought conditions on a coarse temporal scale. Drought indices that describe drought evolution at a fine temporal scale are still scarce. In this study, four machine learning methods, including random forest regression (RFR), k-nearest neighbor regression (KNNR), support vector regression (SVR), and extreme gradient boosting regression (XGBR), were used to construct daily drought indices based on multisource remote sensing and reanalysis data. Through comparison with in situ soil moisture (SM) over the TP, our results indicate that the drought index based on the XGBR model outperforms other models (R2 = 0.76, RMSE = 0.11, MAE = 0.08), followed by RFR (R2 = 0.74, RMSE = 0.11, MAE = 0.08), KNNR (R2 = 0.73, RMSE = 0.11, MAE = 0.08) and SVR (R2 = 0.66, RMSE = 0.12, MAE = 0.1). A new daily drought index, the standardized integrated drought index (SIDI), was developed by the XGBR model for monitoring agricultural drought. A comparison with ERA5-Land SM and widely used indices such as SPI-6 and SPEI-6 indicated that the SIDI depicted the dry and wet change characteristics of the plateau well. Furthermore, the SIDI was applied to analyze a typical drought event and reasonably characterize the spatiotemporal patterns of drought evolution, demonstrating its capability and superiority for drought monitoring over the TP. In addition, soil properties accounted for 59.5% of the model output, followed by meteorological conditions (35.8%) and topographic environment (4.7%). [ABSTRACT FROM AUTHOR]

Soft sensors work as predictive frameworks encapsulating a set of easy-to-collect input data and a machine learning method (ML) to predict highly related variables that are difficult to measure. The machine learning method could provide a prediction of complex unknown relations between the input data and desired output parameters. Recently, soft sensors have been applicable in predicting the prices and vintages of New Zealand Pinot noir wines based on chemical parameters. However, the previous sample size did not adequately represent the diversity of provenances, vintages, and price points across commercially available New Zealand Pinot noir wines. Consequently, a representative sample of 39 commercially available New Zealand Pinot noir wines from diverse provenances, vintages, and price points were selected. Literature has shown that wine phenolic compounds strongly correlated with wine provenances, vintages and price points, which could be used as input data for developing soft sensors. Due to the significance of these phenolic compounds, chemical parameters, including phenolic compounds and pH, were collected using UV-Vis visible spectrophotometry and a pH meter. The soft sensor utilising Naive Bayes (belongs to ML) was designed to predict Pinot noir wines' provenances (regions of origin) based on six chemical parameters with the prediction accuracy of over 75%. Soft sensors based on decision trees (within ML) could predict Pinot noir wines' vintages and price points with prediction accuracies of over 75% based on six chemical parameters. These predictions were based on the same collected six chemical parameters as aforementioned. [ABSTRACT FROM AUTHOR]

Distance learning today allows you to create a system of mass continuous selflearning, universal exchange of information, regardless of the presence of time and space zones. For more effective online learning, it is necessary to introduce artificial intelligence technologies and methods for implementing learning systems. In this paper, the objects of research were: 1) the capabilities of the OSTIS open semantic technology, the capabilities of the IMS.OSTIS metasystem for organizing online learning; 2) machine learning method – classification. Results of the study: 1) an online course was organized using the IMS.OSTIS metasystem of the OSTIS open semantic technology; 2) for the analysis and visualization of training data, a machine learning method is implemented – classification. The results of the implementation of the online course were obtained using the semantic technology for designing an intelligent learning system: the IMS.OSTIS metasystem using the graphical semantic code SCg. The OSTIS kernel requires a machine with the Ubuntu operating system installed, which is a GNU/Linux distribution based on Debian GNU/Linux, an operating system based on the Linux kernel. The paper also shows an example of using the machine learning method – classification. This method allows you to classify data. Intelligent processing and visualization of data were carried out based on the results of testing students in order to classify them into letter categories A, B, C, D according to a set of features: scores and points of the average score. The high-level Python library Pandas was used, this is a library for data analysis. To visualize the results of data processing, the Matplotlib library in Python was used [ABSTRACT FROM AUTHOR]

The Northeast China cold vortex (NCCV) during late summer (from July to August) is identified and classified into three types in terms of its movement path using machine learning. The relationships of the three types of NCCV intensity with atmospheric circulations in late summer, the sea surface temperature (SST), and Arctic sea ice concentration (SIC) in the preceding months, are analyzed. The sensitivity tests by the Community Atmosphere Model version 5.3 (CAM5.3) are used to verify the statistical results. The results show that the coordination pattern of East Asia-Pacific (EAP) and Lake Baikal high pressure forced by SST anomalies in the North Indian Ocean dipole mode (NIOD) during the preceding April and SIC anomalies in the Nansen Basin during the preceding June results in an intensity anomaly for the first type of NCCV. While the pattern of high pressure over the Urals and Okhotsk Sea and low pressure over Lake Baikal during late summer—which is forced by SST anomalies in the South Indian Ocean dipole mode (SIOD) in the preceding June and SIC anomalies in the Barents Sea in the preceding April—causes the intensity anomaly of the second type. The third type is atypical and is not analyzed in detail. Sensitivity tests, jointly forced by the SST and SIC in the preceding period, can well reproduce the observations. In contrast, the results forced separately by the SST and SIC are poor, indicating that the NCCV during late summer is likely influenced by the coordinated effects of both SST and SIC in the preceding months. [ABSTRACT FROM AUTHOR]

The concrete strength criterion is the basis of strength analysis and evaluation under a complex stress state. In this paper, a large number of multiaxial strength tests were carried out, and many mathematical expressions of strength criteria were proposed based on the geometric characteristics and the assumption of a convex function. However, the rationality of the assumption of a convex function limits the use of these strength criteria. In particular, misjudgment will occur near the failure curve surface. Therefore, this paper does not assume the shape function of the criterion in advance. By collecting experimental data and using a machine learning method, it proposes a method of hidden function of failure curve surface. Based on 777 groups of experimental data, the random forest (RF), the back propagation neural network (BP) and the radial basis neural network (RBF) models were used to analyze and verify the feasibility and effectiveness of the method. Subsequently, the results were compared with the Ottosen strength criterion, the Guo Wang strength criterion and the Drucker–Prager (DP) strength criterion. The results show that the consistency between the strength criterion model established by the machine learning algorithm (especially random forest) and the experimental data is higher than the convex function multiaxis strength criterion of the preset failure surface shape. Moreover, the physical significance is clearer, the deficiency of the convex function failure surface hypothesis is avoided and the established multiaxial strength criterion of concrete is more universal. [ABSTRACT FROM AUTHOR]

Nonlinear conservation laws are widely used in fluid mechanics, biology, physics, and chemical engineering. However, deriving such nonlinear conservation laws is a significant and challenging problem. A possible attractive approach is to extract conservation laws more directly from observation data by use of machine learning methods. We propose a framework that combines a symbolic multi-layer neural network and a discrete scheme to learn the nonlinear, unknown flux function f (u) of the scalar conservation law u t + f (u) x = 0 (∗) with u as the main variable. This identification is based on using observation data u (x j , t i) on a spatial grid x j , j = 1 , ... , N x at specified times t i , i = 1 , ... , N o b s . A main challenge with Eq (*) is that the solution typically creates shocks, i.e., one or several jumps of the form (u L , u R) with u L ≠ u R moving in space and possibly changing over time such that information about f (u) in the interval associated with this jump is sparse or not at all present in the observation data. Secondly, the lack of regularity in the solution of (*) and the nonlinear form of f (u) hamper use of previous proposed physics informed neural network (PINN) methods where the underlying form of the sought differential equation is accounted for in the loss function. We circumvent this obstacle by approximating the unknown conservation law (*) by an entropy satisfying discrete scheme where f (u) is represented through a symbolic multi-layer neural network. Numerical experiments show that the proposed method has the ability to uncover the hidden conservation law for a wide variety of different nonlinear flux functions, ranging from pure concave/convex to highly non-convex shapes. This is achieved by relying on a relatively sparse amount of observation data obtained in combination with a selection of different initial data. [ABSTRACT FROM AUTHOR]

Based on the daily minimum temperature data from 199 meteorological stations in northeast China (NEC), the cold wave (CW) processes in winter from 1980 to 2020 are identified. Then, by using the reanalysis 500‐hPa geopotential height data provided by the European Centre for Medium‐Range Weather Forecasts (ECMWF), the CWs are objectively classified into two types (abbreviated as Type 1 and Type 2) according to the corresponding atmospheric circulation through the hierarchical cluster algorithm (HCA). The relationships of the CWs with the Arctic sea sic (SIC) and sea surface temperature (SST) are revealed. For Type 1, the SIC in the north of the Barents Sea–Kara Sea is relatively less in the previous autumn, and a wave train propagating from the polar regions to Northeast Asia is stimulated under the ice–atmosphere interaction. This wave train further cooperates with the positive SST anomaly in the southern Indian Ocean to excite the low‐vortex circulation under the negative phase of Arctic Oscillation (AO) in winter, which is conducive to the occurrence of Type 1. However, the relatively low SST in the western Pacific warm pool in previous autumn stimulates a northward propagating wave train along the meridian, which increases the degree of meridionality of the atmospheric circulation over East Asia and thus indirectly causes the polar‐vortex‐splitting circulation under the positive AO phase, which favours the formation of Type 2. The numerical simulations can well reproduce the observational facts, further confirming the possible impact of the SIC and SST in the previous autumn on the two types of CW over the NEC. [ABSTRACT FROM AUTHOR]

In the classification calculation, the data are sometimes not unique and there are different values and probabilities. Then, it is meaningful to develop the appropriate methods to make classification decision. To solve this issue, this paper proposes the machine learning methods based on a probabilistic decision tree (DT) under the multi-valued preference environment and the probabilistic multi-valued preference environment respectively for the different classification aims. First, this paper develops a data pre-processing method to deal with the weight and quantity matching under the multi-valued preference environment. In this method, we use the least common multiple and weight assignments to balance the probability of each preference. Then, based on the training data, this paper introduces the entropy method to further optimize the DT model under the multi-valued preference environment. After that, the corresponding calculation rules and probability classifications are given. In addition, considering the different numbers and probabilities of the preferences, this paper also uses the entropy method to develop the DT model under the probabilistic multi-valued preference environment. Furthermore, the calculation rules and probability classifications are similarly derived. At last, we demonstrate the feasibility of the machine learning methods and the DT models under the above two preference environments based on the illustrated examples. [ABSTRACT FROM AUTHOR]

Knowledge about tropical cyclone (TC) size is essential for disaster prevention and mitigation strategies, but due to the limitations of observations, TC size data from the open ocean are scarce. In this paper, several models are developed to identify TC size parameters, including the radius of maximum wind (RMW) and the radii of 34 (R34), 50 (R50), and 64 (R64) knot winds, using various machine learning algorithms based on infrared channel imagery of geostationary meteorological satellites over the western North Pacific (WNP). Through evaluation and verification, the trained and optimized support vector machine models are proposed for RMW and R34, whereas the general regression neural network models are set up for R50 and R64. According to the independent-sample evaluations against aircraft observations (1981 - 1987)/Joint Typhoon Warning Center best track data (2017 - 2019), the mean absolute errors of R34, R50, R64, and RMW are 54/58, 34/38, N/A/21, and 25/25 km, respectively. The corresponding median errors are 39/46, 34/31, N/A/17, and 17/19 km, respectively. There is an overall slight underestimation of the parameters, which needs to be analyzed and improved in a future study. Despite aircraft observations of TCs in the WNP having ceased in the late 1980s, this new dataset of TC sizes enables a thorough estimation of wind structures covering a period of 40 years [ABSTRACT FROM AUTHOR]

This paper presents the analysis and intelligent prediction for the displacement of stratum and tunnel lining of Qingdao Metro Line 4 by earth pressure balance (EPB) shield tunnel excavation in complex strata. When the tunnel is excavated in different stratum sections, the tunneling parameters of shield machine are systematically analyzed and compared, and the vertical displacement of the tunnel crown and the horizontal convergence deformation on both sides are investigated. When the tunnel body passes through the soft soil stratum and rock stratum, the curves of the vertical displacement of the stratum surface with time are respectively discussed. A machine learning method for predicting stratum surface deformation induced by shield tunnel excavation in complex strata is developed, where extreme learning machine (ELM), particle swarm optimization (PSO) algorithm and $k$ -fold cross-validation method are comprehensively considered. 65 data samples are collected from the field monitoring data of Qingdao Metro Line 4 and each data sample includes seven input values and one output value. The developed PSO-ELM has good prediction performance for stratum surface vertical displacement due to shield tunnel excavation. The case study in this work can provide a practical reference for similar tunneling projects. [ABSTRACT FROM AUTHOR]

Compared with the attention of large reservoirs, the water quality of small reservoirs also needs attention. In recent years, the problem of reservoir water quality has become increasingly serious. How to predict reservoir water quality may be an urgent problem to be solved. Taking the Yangmeiling reservoir as an example, this paper collects the hydrological and water quality data of the Yangmeiling reservoir in the last ten years, analyzes the relationship between hydrological and water quality data, and uses a machine learning method to simulate the relationship between water quality and hydrological data. The results show that the water quality of small reservoirs can be simply linked with hydrological data and can be predicted through hydrological data, and has high simulation accuracy. This method can be popularized in the simulation and prediction of the water quality of small reservoirs; it does not provide a theoretical basis for the water quality management of small reservoirs. [ABSTRACT FROM AUTHOR]

Traffic classification is considered an important research area due to the increasing demand in network users. It not only effectively improve the network service identifications and security issues of the traffic network, but also provide robust accuracy and efficiency in different Internet application behaviors and patterns. Several traffic classification techniques have been proposed and applied successfully in recent years. However, the existing literature lack of comprehensive survey which could provide an overview and analysis towards the recent developments in network traffic classification. To this end, this survey presents a comprehensive investigation on traffic classification techniques by carefully reviewing existing methods from a new perspective. We comprehensively discuss the procedures and datasets for traffic classification. Additionally, traffic criteria are proposed, which could be beneficial to assess the effectiveness of the developed classification algorithm. Then, the traffic classification techniques are discussed in detail. Then, we thoroughly discussed the machine learning (ML) methods for traffic classification. For researcher’s convenience, we present the traffic obfuscation techniques, which could be helpful for designing a better classifier. Finally, key findings and open research challenges for network traffic classification are identified along with recommendations for future research directions. In sum, this survey fills the gap of existing surveys and summarizes the latest research developments in traffic classification.

With the rapid development of the economy, it is important to reasonably evaluate the development status of the regional manufacturing industry. Given this, this article expands the evaluation indicators of urban advanced manufacturing (UAM) from the perspective of the push–pull-mooring (PPM). Then, it uses a machine learning (ML) method to predict the evaluation results of other cities through a small amount of sample data. The results show that: (1) From the current development status of UAM in Guangdong Province (GD), cities in the Pearl River Delta region occupy a dominant position. However, cities in eastern, western, and mountainous regions have strong development potential and lead cities. Therefore, each region has cities with high levels of development and has a demonstrative role. (2) By comparison, it was found that the overall development level of UAM in GD is not significantly different from that of the Yangtze River Economic Belt. However, due to significant differences in their extreme values, the proportion of cities above the average in the overall population is relatively small. This indirectly proves that GD’s UAM not only has a phased nature, but also has a demonstrative role. (3) The prediction effect of the perceptron model is better than other methods. Although neural network models have better prediction performance than other machine learning models, they should not overly rely on complex network structure prediction data. By comparing the results, the reliability is verified. Finally, according to the life cycle theory, we propose a targeted development path for different UAM.

In this research, we present an efficient implementation of machine learning (ML) models that forecast the mechanical properties of basalt fiber-reinforced high-performance concrete (BFHPC). The objective of the present study was to predict compressive, flexural, and tensile strengths of BFHPC through ML techniques and propose some correlations between these properties. Moreover, the modulus of elasticity (ME) values and compressive stress–strain curves were simulated using ML techniques. In this regard, three predictive algorithms, including linear regression (LR), support vector regression (SVR), and polynomial regression (PR), were considered. LR, SVR, and PR were utilized to forecast the compressive, flexural, and tensile strengths of BFHPC, and the PR technique was employed to simulate the compressive stress–strain curves. The performance of the models was also determined by the coefficient of determination (R2), mean absolute errors (MAE), and root mean square errors (RMSE). According to the obtained values of R2, MAE, and RMSE, the performance of PR was better than other types of algorithms in estimating the compressive, tensile, and flexural strengths. For example, R2 values were 0.99, 0.94, and 0.98 in predicting the compressive, flexural, and tensile strengths using PR, respectively. This shows the higher accuracy and reliability of the PR technique compared with other predictive algorithms. Finally, we concluded that ML techniques can be appropriately applied to assess the mechanical characteristics of BFHPC. [ABSTRACT FROM AUTHOR]

The study of asperities is of great importance for seismic hazard analysis. This paper introduces the concepts of asperities, barriers and their source models, and reviews the progress of research on the determination of asperities by seismic geodesy, mainly discusses the technique and method for determining asperities from the aspects of medium structure analysis （high-intensity anomalies）, seismic activity analysis, the fault slip pattern, and the problems are analyzed. This paper also prospects the determination of the distribution of asperities on active faults, future researches will focus on joint inversion of geodetic data, joint geodetic and seismic geophysical data, and machine learning based methods to determinate the distribution of asperities. The study provides a strong theoretical basis for seismic safety evaluation. [ABSTRACT FROM AUTHOR]

This paper describes the results of a developed online machine learning method implemented in a real-time multiplayer game called Hram Light. The goal was to create a program which produces several ways of playing for a bot, which would become a virtual trainer for players. The solutions are required to vary in success rate in order to achieve progress in player improvement. The input of the method is a data model of parameters which describe a player's way of playing in a game situation. The process of learning takes place in the online world and produces a variety of solutions via a genetic algorithm. [ABSTRACT FROM AUTHOR]

Due to the similar chemical composition and matrix effect, the accurate identification of mineral pigments on wall paintings has brought great challenges. This work implemented an identification study on three mineral pigments with similar chemical compositions by combining LIBS technology with the K-nearest neighbor algorithm (KNN), random forest (RF support vector machine (SVM), back propagation artificial neural network (Bp-ANN) and convolutional neural network (CNN) to find the most suitable identification method for mural research. Using the SelectKBest algorithm, 300 characteristic lines with the largest difference among the three pigments were determined. The identification models of KNN, RF, SVM, Bp-ANN and CNN were established and optimized. The results showed that, except for the KNN model, the identification accuracy of other models for mock-up mural samples was above 99%. However, only the identification accuracy of 2D-CNN models reached above 94% for actual mural samples. Therefore, the 2D-CNN model was determined as the most suitable model for the identification and analysis of mural pigments. [ABSTRACT FROM AUTHOR]

Designing plate structures poses challenges due to potential instabilities and the complexities involved in obtaining accurate analytical solutions, especially for non-Lévy-type boundary conditions. This study employs the finite integral transform method to analyze the buckling behavior of orthotropic thin plates with complex boundary conditions. In solution procedure, the governing high-order partial differential equations are transformed into a system of linear algebraic equations, yielding exact and rapidly converging analytical solution. The method is simple and general and does not need to predetermine the deflection function. The correctness of the method is validated by comparison with numerical simulations performed using the ABAQUS software. Furthermore, Gene Expression Programming (GEP) is utilized to develop empirical models that predict buckling coefficients of isotropic and orthotropic plates under classical and non-classical boundary conditions. Material properties, aspect ratio, rotating fixed coefficient, and boundary conditions are used as input parameters, with simplified mathematical formulations to predict the buckling coefficient. Model performance is assessed using parametric analysis and statistical tests to ensure accuracy and generalization. Further investigation shows aspect ratio and rotating fixed coefficient are significant variables influencing buckling coefficients under classical and non-classical boundary conditions, respectively, followed by boundary conditions and material property. The performance of the GEP model is further assessed by comparing with linear and non-linear regression models. The results show that GEP outperforms the regression models, showing its higher prediction accuracy. This study not only addresses the challenges of designing thin plates with complex boundary conditions, but it also presents an effective machine-learning method for predicting buckling behavior. The analytical solution can be used as a benchmark for validating new analytical and numerical methods. The equations developed using GEP to predict the buckling coefficient of plates provide a useful tool for extrapolating results beyond the scope of this study. [Display omitted] • A novel finite integral transform method is employed to investigate the buckling behavior of thin plates. • Analytical solution is obtained for thin plates with classical and non-classical types of boundary conditions. • Gene Expression Programming is used to accurately predict the buckling coefficients of thin plates under different scenarios. [ABSTRACT FROM AUTHOR]

The value of a firm is important for shareholders, investors, managers, creditors and other stakeholders in order to assess its future and to estimate the risk and return on investment and stock prices. Investors have always concerns about estimating the future value of firms and they use various financial instruments in this regard. The purpose of this research is to explain the value of a firm using performance, management system and audit committee variables and applying the Lars machine learning method, as well as to forecast the value of the firm using the Lasso linear and Gaussian process nonlinear methods to help decision makers and investors. For this purpose, data is collected from 208 years/firms listed in Tehran Stock Exchange during the seven financial period of 2011-2017. The preliminary results of the research show that performance criteria are more capable of explaining the value of a firm than the group of criteria of the management system and the audit committee. Also other results of the research indicate the high power of machine learning methods for forecasting the value of a firm, in particular, the Gaussian process nonlinear approach as compared to the linear method of the Lasso is.

As the two important ambient air pollutants, particulate matter (PM 2.5) and ozone (O 3) can both originate from gas nitrogen oxides. In this study, applied by theoretical analysis and machine learning method, we examined the effects of atmospheric reactive nitrogen on PM 2.5 -O 3 pollution, in which nitric oxide (NO), nitrogen dioxide (NO 2), gaseous nitric acid (HNO 3) and particle nitrate (pNO 3 −) conversion process has the co-directional and contra-directional effects on PM 2.5 -O 3 pollution. Of which, HNO 3 and SO 2 are the co-directional driving factors resulting in PM 2.5 and O 3 growing or decreasing simultaneously; while NO, NO 2 , and temperature represent the contra-directional factors, which can promote the growth of one pollutant and reduce another one. Our findings suggest that designing the suitable co-controlling strategies for PM 2.5 -O 3 sustainable reduction should target at driving factors by considering the contra-directional and co-directional effects under suitable sensitivity regions. For co-directional driving factors, the design of suitable mitigation strategies will jointly achieve effective reduction in PM 2.5 and O 3 ; while for contra-directional driving factors, it should be more patient, otherwise, it is possible to reduce one item but increase another one at the same time. [Display omitted] • Atmospheric reactive nitrogen conversion regulates PM 2.5 -O 3 pollution. • The co-directional and contra-directional effects on PM 2.5 -O 3 pollution. • HNO 3 and SO 2 facilitate PM 2.5 and O 3 growing or decreasing, simultaneously. • NO x and temperature can promote the growth of one pollutant and reduce another one. [ABSTRACT FROM AUTHOR]

The environmental situation in the capital city is always in the focus of attention of the municipal authorities of the city and is one of the most important factors influencing the decisions. The capital of Kazakhstan, Nur-Sultan, consumes heat energy generated from fossil fuel, and one of the major problems is an extremely cold and long winter. The GHG emissions and particle matters from the coal-based Combined Heat and Power plant have a significant impact on the environment as smog, particularly in the heating season. This work analyses spatial high-resolution Big Data collected from the metering points of 385 houses and 62 heat transmission contours across a city during the heating season. The temporary resolution was 10 minutes i.e., 8754 rows for 5 months (Jan-May). There are shown the findings of the correlation rates analysis between heat energy consumption by zones of Nur-Sultan and ambient temperature, as well as non-efficient zones with substantial losses. In this paper for developing the prediction tools for the Smart City heat consumption there were used mixed modelling methods and machine learning approaches, such as Linear Regression, Kneighbours Regressor, and Random Forest Regressor models. These obtained results could be helpful for predicting optimal pumping pressure for each distribution network using machine learning technologies and finding overheated contours in real time. [ABSTRACT FROM AUTHOR]

This paper deals with the prediction of the future location of vehicles, which is attracting attention in the era of the fourth industrial revolution and is required in various fields, such as autonomous vehicles and smart city traffic management systems. Currently, vehicle traffic prediction models and accident prediction models are being tested in various places, and considerable progress is being made. However, there are always errors in positioning when using wireless sensors due to various variables, such as the appearance of various substances (water, metal) that occur in the space where radio waves exist. There have been various attempts to reduce the positioning error in such an Internet of Things environment, but there is no definitive method with confirmed performance. Of course, location prediction is also not accurate. In particular, since a vehicle moves rapidly in space, it is increasingly affected by changes in the environment. Firstly, it was necessary to develop a spatial positioning algorithm that can improve the positioning accuracy. Secondly, for the data generated by the positioning algorithm, a machine learning method suitable for position prediction was developed. Based on the above two developed algorithms, through experiments, we found a means to reduce the error of positioning through radio waves and to increase the accuracy of positioning. We started with the idea of changing the positioning space itself from a three-dimensional space into a two-dimensional one. With changes in the time and space of radio wave measurement, the location was measured by transforming the spatial dimension to cope with environmental changes. This is a technology that predicts a location through machine learning on time series data using a direction angle classification technique. An experiment was conducted to verify the performance of the proposed technology. As a result, the accuracy of positioning was improved, and the accuracy of location prediction increased in proportion to the learning time. It was possible to confirm the prediction accuracy increase of up to 80% with changes. Considering that the accuracy result for location prediction presented by other researchers is 70%, through this study, the result was improved by 10% compared to the existing vehicle location prediction accuracy. In conclusion, this paper presents a positioning algorithm and machine learning methodology for vehicle positioning. By proving its usefulness through experiments, this study provides other researchers with a new definition of space for predicting the location of a vehicle, and a machine learning method using direction angles. [ABSTRACT FROM AUTHOR]

Remaining useful life prediction is an important way to improve the durability and reduce the cost of the proton exchange membrane fuel cell. This paper presents a novel method to predict the remaining useful life of proton exchange membrane fuel cell under different load currents based on support vector regression and grey wolf optimizer algorithm. The proposed method considers the influence of 17 operating conditions and historical voltage. Firstly, the measured data are reconstructed through robust locally weighted smoothing method to reduce the calculation amount and filter disturbances. Then, support vector regression with fewer hyperparameters is used to establish the degradation model. Finally, the hyperparameters of support vector regression are optimized through grey wolf optimizer algorithm to improve the accuracy of degradation prediction. The proposed method is validated by two degradation experiments under different load currents. The test results show that grey wolf optimizer algorithm can effectively improve the accuracy of degradation prediction based on support vector regression. Compared with other methods, the proposed method has the highest accuracy. The proposed method can predict the fuel cell degradation with a mean absolute percentage error of less than 0.3%. The proposed method can predict the remaining useful life of 492 hours. [ABSTRACT FROM AUTHOR]

Abstract Background Self-esteem is the individual evaluation of oneself. People with high self-esteem grade have mental health and can bravely cope with the threats from the environment. With the development of neuroimaging techniques, researches on cognitive neural mechanisms of self-esteem are increased. Existing methods based on brain morphometry and single-layer brain network cannot characterize the subtle structural differences related to self-esteem. Method To solve this issue, we proposed a multiple anatomical brain network based on multi-resolution region of interest (ROI) template to study the brain structural connections of self-esteem. The multiple anatomical brain network consists of ROI features and hierarchal brain network features that are extracted from structural MRI. For each layer, we calculated the correlation relationship between pairs of ROIs. In order to solve the high-dimensional problem caused by the large amount of network features, feature selection methods (t-test, mRMR, and SVM-RFE) are adopted to reduce the number of features while retaining discriminative information to the maximum extent. Multi-kernel SVM is employed to integrate the various types of features by appropriate weight coefficient. Result The experimental results show that the proposed method can improve classification accuracy to 97.26% compared with single-layer brain network. Conclusions The proposed method provides a new perspective for the analysis of brain structural differences of self-esteem, which also has potential guiding significance in other researches involved brain cognitive activity and brain disease diagnosis.

There is a growing attention to the bio and renewable energies due to fast depletion of fossil fuels as well as the global warming problem. Here, we developed a modeling and simulation method by means of artificial intelligence (AI) for prediction of the bioenergy production from vegetable bean oil. AI methods are well known for prediction of complex and nonlinear process. Three distinct Adaptive Boosted models including Huber regression, LASSO, and Support Vector Regression (SVR) as well as artificial neural network (ANN) were applied in this study to predict actual yield of Fatty acid methyl esters (FAME) production. All boosted utilizing the Adaptive boosting algorithm. The important influencing parameters on the biodiesel production such as the catalyst loading (CAO/Ag, wt%) and methanol to oil (Soybean oil) molar ratio were selected as the input variables of models while the yield of FAME production was selected as output. Model hyper-parameters were tuned to maintain generality while improving prediction accuracy. The models were evaluated using three distinct metrics Mean Absolute Error (MAE), Root Mean Square Error (RMSE), and R2. Error rates of 8.16780E-01, 4.43895E-01, 2.06692E + 00, and 3.92713 E-01 were obtained with the MAE metric for boosted Huber, SVR, LASSO and ANN models. On the other hand, the RMSE error of these models were about 1.092E-02, 1.015E-02, 2.669E-02, and 1.01174E-02, respectively. Finally, the R-square score were calculated for boosted Huber, boosted SVR, and boosted LASSO as 0.976, 0.990, 0.872, and 0.99702, respectively. Therefore, it can be concluded that although the boosted SVR and ANN models were better models for prediction of process efficiency in terms of error, but all algorithms had high accuracy. The optimum yield of 83.77% and 81.60% for biodiesel production were observed at optimum operating values from boosted SVR and ANN models, respectively.