Deep learning for determining pure isotropic proton spectra from solidstate spectra

Recently,an article on ^(1)H solid-state NMR spectra was published,in which the authors proposed a deep learning approach to infer the pure isotropic proton NMR spectra obtained at an infinite magic angle spinning(MAS)rate.This approach even allowed to obtain,by far,the best resolved ^(1)H spectra of molecular solids[1](https://doi.org/10.1002/anie.202216607).Deep learning based artificial intelligence is developing rapidly,and its application is deepening.Currently,there are many applications of deep learning in the field of magnetic resonance,such as the reconstruction of the under-sampled multidimensional spectra[2-4],the deconvolution of two-dimensional NMR spectra[5]and noise suppression and weak peak retrial[6],etc.

Deep Learning for Financial Time Series Prediction:A State-of-the-Art Review of Standalone and HybridModels

Financial time series prediction,whether for classification or regression,has been a heated research topic over the last decade.While traditional machine learning algorithms have experienced mediocre results,deep learning has largely contributed to the elevation of the prediction performance.Currently,the most up-to-date review of advanced machine learning techniques for financial time series prediction is still lacking,making it challenging for finance domain experts and relevant practitioners to determine which model potentially performs better,what techniques and components are involved,and how themodel can be designed and implemented.This review article provides an overview of techniques,components and frameworks for financial time series prediction,with an emphasis on state-of-the-art deep learning models in the literature from2015 to 2023,including standalonemodels like convolutional neural networks(CNN)that are capable of extracting spatial dependencies within data,and long short-term memory(LSTM)that is designed for handling temporal dependencies;and hybrid models integrating CNN,LSTM,attention mechanism(AM)and other techniques.For illustration and comparison purposes,models proposed in recent studies are mapped to relevant elements of a generalized framework comprised of input,output,feature extraction,prediction,and related processes.Among the state-of-the-artmodels,hybrid models like CNNLSTMand CNN-LSTM-AM in general have been reported superior in performance to stand-alone models like the CNN-only model.Some remaining challenges have been discussed,including non-friendliness for finance domain experts,delayed prediction,domain knowledge negligence,lack of standards,and inability of real-time and highfrequency predictions.The principal contributions of this paper are to provide a one-stop guide for both academia and industry to review,compare and summarize technologies and recent advances in this area,to facilitate smooth and informed implementation,and to highlight future research directions.

Tool Wear State Recognition with Deep Transfer Learning Based on Spindle Vibration for Milling Process

The wear of metal cutting tools will progressively rise as the cutting time goes on. Wearing heavily on the toolwill generate significant noise and vibration, negatively impacting the accuracy of the forming and the surfaceintegrity of the workpiece. Hence, during the cutting process, it is imperative to continually monitor the tool wearstate andpromptly replace anyheavilyworn tools toguarantee thequality of the cutting.The conventional tool wearmonitoring models, which are based on machine learning, are specifically built for the intended cutting conditions.However, these models require retraining when the cutting conditions undergo any changes. This method has noapplication value if the cutting conditions frequently change. This manuscript proposes a method for monitoringtool wear basedonunsuperviseddeep transfer learning. Due to the similarity of the tool wear process under varyingworking conditions, a tool wear recognitionmodel that can adapt to both current and previous working conditionshas been developed by utilizing cutting monitoring data from history. To extract and classify cutting vibrationsignals, the unsupervised deep transfer learning network comprises a one-dimensional (1D) convolutional neuralnetwork (CNN) with a multi-layer perceptron (MLP). To achieve distribution alignment of deep features throughthe maximum mean discrepancy algorithm, a domain adaptive layer is embedded in the penultimate layer of thenetwork. A platformformonitoring tool wear during endmilling has been constructed. The proposedmethod wasverified through the execution of a full life test of end milling under multiple working conditions with a Cr12MoVsteel workpiece. Our experiments demonstrate that the transfer learning model maintains a classification accuracyof over 80%. In comparisonwith the most advanced tool wearmonitoring methods, the presentedmodel guaranteessuperior performance in the target domains.

Optimal Classification of Minerals by Microscopic Image Analysis Based on Seven-State “Deep Learning” Combined with Optimizers

The development of artificial intelligence (AI), particularly deep learning, has made it possible to accelerate and improve the processing of data collected in different fields (commerce, medicine, surveillance or security, agriculture, etc.). Most related works use open source consistent image databases. This is the case for ImageNet reference data such as coco data, IP102, CIFAR-10, STL-10 and many others with variability representatives. The consistency of its images contributes to the spectacular results observed in its fields with deep learning. The application of deep learning which is making its debut in geology does not, to our knowledge, include a database of microscopic images of thin sections of open source rock minerals. In this paper, we evaluate three optimizers under the AlexNet architecture to check whether our acquired mineral images have object features or patterns that are clear and distinct to be extracted by a neural network. These are thin sections of magmatic rocks (biotite and 2-mica granite, granodiorite, simple granite, dolerite, charnokite and gabbros, etc.) which served as support. We use two hyper-parameters: the number of epochs to perform complete rounds on the entire data set and the “learning rate” to indicate how quickly the weights in the network will be modified during optimization. Using Transfer Learning, the three (3) optimizers all based on the gradient descent methods of Stochastic Momentum Gradient Descent (sgdm), Root Mean Square Propagation (RMSprop) algorithm and Adaptive Estimation of moment (Adam) achieved better performance. The recorded results indicate that the Momentum optimizer achieved the best scores respectively of 96.2% with a learning step set to 10−3 for a fixed choice of 350 epochs during this variation and 96, 7% over 300 epochs for the same value of the learning step. This performance is expected to provide excellent insight into image quality for future studies. Then they participate in the development of an intelligent system for the identification and classification of minerals, seven (7) in total (quartz, biotite, amphibole, plagioclase, feldspar, muscovite, pyroxene) and rocks.

Early identification of stroke through deep learning with multi-modal human speech and movement data

Early identification and treatment of stroke can greatly improve patient outcomes and quality of life.Although clinical tests such as the Cincinnati Pre-hospital Stroke Scale(CPSS)and the Face Arm Speech Test(FAST)are commonly used for stroke screening,accurate administration is dependent on specialized training.In this study,we proposed a novel multimodal deep learning approach,based on the FAST,for assessing suspected stroke patients exhibiting symptoms such as limb weakness,facial paresis,and speech disorders in acute settings.We collected a dataset comprising videos and audio recordings of emergency room patients performing designated limb movements,facial expressions,and speech tests based on the FAST.We compared the constructed deep learning model,which was designed to process multi-modal datasets,with six prior models that achieved good action classification performance,including the I3D,SlowFast,X3D,TPN,TimeSformer,and MViT.We found that the findings of our deep learning model had a higher clinical value compared with the other approaches.Moreover,the multi-modal model outperformed its single-module variants,highlighting the benefit of utilizing multiple types of patient data,such as action videos and speech audio.These results indicate that a multi-modal deep learning model combined with the FAST could greatly improve the accuracy and sensitivity of early stroke identification of stroke,thus providing a practical and powerful tool for assessing stroke patients in an emergency clinical setting.

Deep eutectic solvents for separation and purification applications in critical metal metallurgy:Recent advances and perspectives

Solvent extraction,a separation and purification technology,is crucial in critical metal metallurgy.Organic solvents commonly used in solvent extraction exhibit disadvantages,such as high volatility,high toxicity,and flammability,causing a spectrum of hazards to human health and environmental safety.Neoteric solvents have been recognized as potential alternatives to these harmful organic solvents.In the past two decades,several neoteric solvents have been proposed,including ionic liquids(ILs)and deep eutectic solvents(DESs).DESs have gradually become the focus of green solvents owing to several advantages,namely,low toxicity,degradability,and low cost.In this critical review,their classification,formation mechanisms,preparation methods,characterization technologies,and special physicochemical properties based on the most recent advancements in research have been systematically described.Subsequently,the major separation and purification applications of DESs in critical metal metallurgy were comprehensively summarized.Finally,future opportunities and challenges of DESs were explored in the current research area.In conclusion,this review provides valuable insights for improving our overall understanding of DESs,and it holds important potential for expanding separation and purification applications in critical metal metallurgy.

An Enhanced Lung Cancer Detection Approach Using Dual-Model Deep Learning Technique

Lung cancer continues to be a leading cause of cancer-related deaths worldwide,emphasizing the critical need for improved diagnostic techniques.Early detection of lung tumors significantly increases the chances of successful treatment and survival.However,current diagnostic methods often fail to detect tumors at an early stage or to accurately pinpoint their location within the lung tissue.Single-model deep learning technologies for lung cancer detection,while beneficial,cannot capture the full range of features present in medical imaging data,leading to incomplete or inaccurate detection.Furthermore,it may not be robust enough to handle the wide variability in medical images due to different imaging conditions,patient anatomy,and tumor characteristics.To overcome these disadvantages,dual-model or multi-model approaches can be employed.This research focuses on enhancing the detection of lung cancer by utilizing a combination of two learning models:a Convolutional Neural Network(CNN)for categorization and the You Only Look Once(YOLOv8)architecture for real-time identification and pinpointing of tumors.CNNs automatically learn to extract hierarchical features from raw image data,capturing patterns such as edges,textures,and complex structures that are crucial for identifying lung cancer.YOLOv8 incorporates multiscale feature extraction,enabling the detection of tumors of varying sizes and scales within a single image.This is particularly beneficial for identifying small or irregularly shaped tumors that may be challenging to detect.Furthermore,through the utilization of cutting-edge data augmentation methods,such as Deep Convolutional Generative Adversarial Networks(DCGAN),the suggested approach can handle the issue of limited data and boost the models’ability to learn from diverse and comprehensive datasets.The combined method not only improved accuracy and localization but also ensured efficient real-time processing,which is crucial for practical clinical applications.The CNN achieved an accuracy of 97.67%in classifying lung tissues into healthy and cancerous categories.The YOLOv8 model achieved an Intersection over Union(IoU)score of 0.85 for tumor localization,reflecting high precision in detecting and marking tumor boundaries within the images.Finally,the incorporation of synthetic images generated by DCGAN led to a 10%improvement in both the CNN classification accuracy and YOLOv8 detection performance.

High-power,narrow linewidth solid-state deep ultraviolet laser generation at 193 nm by frequency mixing in LBO crystals

A 60-mW solid-state deep ultraviolet(DUV)laser at 193 nm with narrow linewidth is obtained with two stages of sum frequency generation in LBO crystals.The pump lasers,at 258 and 1553 nm,are derived from a homemade Yb-hybrid laser employing fourth-harmonic generation and Er-doped fiber laser,respectively.The Yb-hybrid laser,finally,is power scaling by a 2 mm×2 mm×30 mm Yb:YAG bulk crystal.Accompanied by the generated 220-mW DUV laser at 221 nm,the 193-nm laser delivers an average power of 60 mW with a pulse duration of 4.6 ns,a repetition rate of 6 kHz,and a linewidth of∼640 MHz.To the best of our knowledge,this is the highest power of 193-and 221-nm laser generated by an LBO crystal ever reported as well as the narrowest linewidth of 193-nm laser by it.Remarkably,the conversion efficiency reaches 27%for 221 to 193 nm and 3%for 258 to 193 nm,which are the highest efficiency values reported to date.We demonstrate the huge potential of LBO crystals for producing hundreds of milliwatt or even watt level 193-nm laser,which also paves a brand-new way to generate other DUV laser wavelengths.

Comparative analysis of empirical and deep learning models for ionospheric sporadic E layer prediction

Sporadic E(Es)layers in the ionosphere are characterized by intense plasma irregularities in the E region at altitudes of 90-130 km.Because they can significantly influence radio communications and navigation systems,accurate forecasting of Es layers is crucial for ensuring the precision and dependability of navigation satellite systems.In this study,we present Es predictions made by an empirical model and by a deep learning model,and analyze their differences comprehensively by comparing the model predictions to satellite RO measurements and ground-based ionosonde observations.The deep learning model exhibited significantly better performance,as indicated by its high coefficient of correlation(r=0.87)with RO observations and predictions,than did the empirical model(r=0.53).This study highlights the importance of integrating artificial intelligence technology into ionosphere modelling generally,and into predicting Es layer occurrences and characteristics,in particular.

Advancements in Liver Tumor Detection:A Comprehensive Review of Various Deep Learning Models

Liver cancer remains a leading cause of mortality worldwide,and precise diagnostic tools are essential for effective treatment planning.Liver Tumors(LTs)vary significantly in size,shape,and location,and can present with tissues of similar intensities,making automatically segmenting and classifying LTs from abdominal tomography images crucial and challenging.This review examines recent advancements in Liver Segmentation(LS)and Tumor Segmentation(TS)algorithms,highlighting their strengths and limitations regarding precision,automation,and resilience.Performance metrics are utilized to assess key detection algorithms and analytical methods,emphasizing their effectiveness and relevance in clinical contexts.The review also addresses ongoing challenges in liver tumor segmentation and identification,such as managing high variability in patient data and ensuring robustness across different imaging conditions.It suggests directions for future research,with insights into technological advancements that can enhance surgical planning and diagnostic accuracy by comparing popular methods.This paper contributes to a comprehensive understanding of current liver tumor detection techniques,provides a roadmap for future innovations,and improves diagnostic and therapeutic outcomes for liver cancer by integrating recent progress with remaining challenges.

Prediction of microstructure evolution of ZK61 alloy during hot spinning by internal state variable model

An internal state variable(ISV)model was established according to the experimental results of hot plane strain compression(PSC)to predict the microstructure evolution during hot spinning of ZK61 alloy.The effects of the internal variables were considered in this ISV model,and the parameters were optimized by genetic algorithm.After validation,the ISV model was used to simulate the evolution of grain size(GS)and dynamic recrystallization(DRX)fraction during hot spinning via Abaqus and its subroutine Vumat.By comparing the simulated results with the experimental results,the application of the ISV model was proven to be reliable.Meanwhile,the strength of the thin-walled spun ZK61 tube increased from 303 to 334 MPa due to grain refinement by DRX and texture strengthening.Besides,some ultrafine grains(0.5μm)that played an important role in mechanical properties were formed due to the proliferation,movement,and entanglement of dislocations during the spinning process.

Machine learning and deep learning to improve prevention of anastomotic leak after rectal cancer surgery

Anastomotic leakage(AL)is a significant complication following rectal cancer surgery,adversely affecting both quality of life and oncological outcomes.Recent advancements in artificial intelligence(AI),particularly machine learning and deep learning,offer promising avenues for predicting and preventing AL.These technologies can analyze extensive clinical datasets to identify preoperative and perioperative risk factors such as malnutrition,body composition,and radiological features.AI-based models have demonstrated superior predictive power compared to traditional statistical methods,potentially guiding clinical decisionmaking and improving patient outcomes.Additionally,AI can provide surgeons with intraoperative feedback on blood supply and anatomical dissection planes,minimizing the risk of intraoperative complications and reducing the likelihood of AL development.

Physically Constrained Adaptive Deep Learning for Ocean Vertical-Mixing Parameterization

Existing traditional ocean vertical-mixing schemes are empirically developed without a thorough understanding of the physical processes involved,resulting in a discrepancy between the parameterization and forecast results.The uncertainty in ocean-mixing parameterization is primarily responsible for the bias in ocean models.Benefiting from deep-learning technology,we design the Adaptive Fully Connected Module with an Inception module as the baseline to minimize bias.It adaptively extracts the best features through fully connected layers with different widths,and better learns the nonlinear relationship between input variables and parameterization fields.Moreover,to obtain more accurate results,we impose KPP(K-Profile Parameterization)and PP(Pacanowski–Philander)schemes as physical constraints to make the network parameterization process follow the basic physical laws more closely.Since model data are calculated with human experience,lacking some unknown physical processes,which may differ from the actual data,we use a decade-long time record of hydrological and turbulence observations in the tropical Pacific Ocean as training data.Combining physical constraints and a nonlinear activation function,our method catches its nonlinear change and better adapts to the oceanmixing parameterization process.The use of physical constraints can improve the final results.

How Do Deep Learning Forecasting Models Perform for Surface Variables in the South China Sea Compared to Operational Oceanography Forecasting Systems?

It is fundamental and useful to investigate how deep learning forecasting models(DLMs)perform compared to operational oceanography forecast systems(OFSs).However,few studies have intercompared their performances using an identical reference.In this study,three physically reasonable DLMs are implemented for the forecasting of the sea surface temperature(SST),sea level anomaly(SLA),and sea surface velocity in the South China Sea.The DLMs are validated against both the testing dataset and the“OceanPredict”Class 4 dataset.Results show that the DLMs'RMSEs against the latter increase by 44%,245%,302%,and 109%for SST,SLA,current speed,and direction,respectively,compared to those against the former.Therefore,different references have significant influences on the validation,and it is necessary to use an identical and independent reference to intercompare the DLMs and OFSs.Against the Class 4 dataset,the DLMs present significantly better performance for SLA than the OFSs,and slightly better performances for other variables.The error patterns of the DLMs and OFSs show a high degree of similarity,which is reasonable from the viewpoint of predictability,facilitating further applications of the DLMs.For extreme events,the DLMs and OFSs both present large but similar forecast errors for SLA and current speed,while the DLMs are likely to give larger errors for SST and current direction.This study provides an evaluation of the forecast skills of commonly used DLMs and provides an example to objectively intercompare different DLMs.

Machine learning based damage state identification:A novel perspective on fragility analysis for nuclear power plants considering structural uncertainties

Seismic fragility analysis(SFA)is known as an effective probabilistic-based approach used to evaluate seismic fragility.There are various sources of uncertainties associated with this approach.A nuclear power plant(NPP)system is an extremely important infrastructure and contains many structural uncertainties due to construction issues or structural deterioration during service.Simulation of structural uncertainties effects is a costly and time-consuming endeavor.A novel approach to SFA for the NPP considering structural uncertainties based on the damage state is proposed and examined.The results suggest that considering the structural uncertainties is essential in assessing the fragility of the NPP structure,and the impact of structural uncertainties tends to increase with the state of damage.Subsequently,machine learning(ML)is found to be superior in high-precision damage state identification of the NPP for reducing the time of nonlinear time-history analysis(NLTHA)and could be applied in the damage state-based SFA.Also,the impact of various sources of uncertainties is investigated through sensitivity analysis.The Sobol and Shapley additive explanations(SHAP)method can be complementary to each other and able to solve the problem of quantifying seismic and structural uncertainties simultaneously and the interaction effect of each parameter.

Optimizing crop yields while minimizing environmental impact through deep placement of nitrogen fertilizer

Nitrogen(N)serves as an essential nutrient for yield formation across diverse crop types.However,agricultural production encounters numerous challenges,notably high N fertilizer rates coupled with low N use efficiency and serious environmental pollution.Deep placement of nitrogen fertilizer(DPNF)is an agronomic measure that shows promise in addressing these issues.This review aims to offer a comprehensive understanding of DPNF,beginning with a succinct overview of its development and methodologies for implementation.Subsequently,the optimal fertilization depth and influencing factors for different crops are analyzed and discussed.Additionally,it investigates the regulation and mechanism underlying the DPNF on crop development,yield,N use efficiency and greenhouse gas emissions.Finally,the review delineates the limitations and challenges of this technology and provides suggestions for its improvement and application.This review provides valuable insight and reference for the promotion and adoption of DPNF in agricultural practice.

基于Deep Forest算法的对虾急性肝胰腺坏死病(AHPND)预警数学模型构建

为预报池塘养殖凡纳对虾(Penaeus vannamei)急性肝胰腺坏死病(AHPND)的发生,自2020年开始,笔者对凡纳对虾养殖区开展了连续监测工作,包括与疾病发生相关的环境理化因子、微生物因子、虾体自身健康状况等18个候选预警因子指标,通过数据标准化处理后分析病原、宿主与环境之间的相关性,对候选预警因子进行筛选,基于Python语言编程结合Deep Forest、Light GBM、XGBoost算法进行数据建模和预测性能评判,仿真环境为Python2.7,以预警因子指标作为输入样本(即警兆),以对虾是否发病指标作为输出结果(即警情),根据输入样本和输出结果各自建立输入数据矩阵和目标数据矩阵,利用原始数据矩阵对输入样本进行初始化,结合函数方程进行拟合,拟合的源代码能利用已知环境、病原及对虾免疫指标数据对目标警情进行预测。最终建立了基于Deep Forest算法的虾体(肝胰腺内)细菌总数、虾体弧菌(Vibrio)占比、水体细菌总数和盐度的4维向量预警预报模型,准确率达89.00%。本研究将人工智能算法应用到对虾AHPND发生的预测预报,相关研究结果为对虾AHPND疾病预警预报建立了预警数学模型,并为对虾健康养殖和疾病防控提供了技术支撑和有力保障。

2-D elastic FEM simulation on stress state in the deep part of a subducted slab

Based upon some simplified numerical models, a 2-D plain strain elastic FEM program is compiled to study the distributions of the stress fields produced by the volume change of the phase transformation from olivine to spinel, by the volume change from temperature variation, and by density difference and boundary action in a piece of subducted slab located in transition zone of the mantle. Thermal stress could explain the fault plane solutions of deep focus earthquakes, but could not explain the distribution of deep seismicity. When large extent metastable olivine is included, the stress field produced by the density difference contradicts with the results of fault plane solutions and with the distribution of deep seismicity. Although the stress produced by volume change of the phase transformation from olivine to spinel dominates the stress state, its main direction is different from the observed results. We conclude that the deep seismicity could not be simply explained by elastic simulation.

A Sensorless State Estimation for A Safety-Oriented Cyber-Physical System in Urban Driving:Deep Learning Approach

In today's modern electric vehicles,enhancing the safety-critical cyber-physical system(CPS)'s performance is necessary for the safe maneuverability of the vehicle.As a typical CPS,the braking system is crucial for the vehicle design and safe control.However,precise state estimation of the brake pressure is desired to perform safe driving with a high degree of autonomy.In this paper,a sensorless state estimation technique of the vehicle's brake pressure is developed using a deep-learning approach.A deep neural network(DNN)is structured and trained using deep-learning training techniques,such as,dropout and rectified units.These techniques are utilized to obtain more accurate model for brake pressure state estimation applications.The proposed model is trained using real experimental training data which were collected via conducting real vehicle testing.The vehicle was attached to a chassis dynamometer while the brake pressure data were collected under random driving cycles.Based on these experimental data,the DNN is trained and the performance of the proposed state estimation approach is validated accordingly.The results demonstrate high-accuracy brake pressure state estimation with RMSE of 0.048 MPa.

A review of deep learning approach to predicting the state of health and state of charge of lithium-ion batteries

In the field of energy storage,it is very important to predict the state of charge and the state of health of lithium-ion batteries.In this paper,we review the current widely used equivalent circuit and electrochemical models for battery state predictions.The review demonstrates that machine learning and deep learning approaches can be used to construct fast and accurate data-driven models for the prediction of battery performance.The details,advantages,and limitations of these approaches are presented,compared,and summarized.Finally,future key challenges and opportunities are discussed.