Search

Handwriting is a biometric behavioral characteristic with evident individual distinctiveness. With the rise of the deep learning trend and demands for forensic identification, handwriting identification has become one of the focal points of research in the field of pattern recognition. Research in handwriting identification for major global languages has matured. However, in China, there is limited attention in the field of writer identification for minority languages such as Mongolian, making it challenging to resolve criminal cases involving handwriting issues. This paper initiates an initial exploration of Mongolian handwriting identification by constructing a structurally simple convolutional neural network. This convolutional neural network, consisting of 12 convolution operations and designed for Mongolian handwriting identification, is referred to as MWInet-12. In this paper, the model evaluation experiments were conducted using a dataset comprising 156,372 samples contributed by 125 writers from the MOLHW dataset. The dataset was divided into training, validation, and test sets in an 8:1:1 ratio. The final results of the experiments reveal impressive accuracy on the test set, achieving a top-1 accuracy of 89.60% and a top-5 accuracy of 97.53%. Furthermore, through comparative experiments involving Resnet50, Fragnet, GRRNN, VGG16, and VGG19 models, this paper establishes that the proposed model yields the most favorable results for Mongolian handwriting identification. The exploratory research on Mongolian handwriting identification in this paper contributes to increasing awareness of information processing for minority languages. It aids in advancing research on classifying writers of Mongolian historical texts and provides technical support for judicial authentication involving handwriting issues. [ABSTRACT FROM AUTHOR]

COVID-19 is one of the leading causes of death worldwide in the year 2020 and was declared a pandemic by the World Health Organization (WHO). This virus affects all countries across the world and 5 lakh people died as of June 2020 due to COVID-19. Due to the highly contagious nature, early detection of this virus plays a vital role to break the covid chain. Recent studies done by China say that chest CT and x-ray image may be used as a preliminary test for covid detection. Deep learning-based CNN model can detect coronavirus automatically from the chest x-ray images. This paper proposed a transfer learning-based approach to detect covid disease. Due to the low number of covid chest images, the authors are using a pre-trained model to classify x-ray images into covid and normal classes. This paper presents the comparative study of various pre-trained models like VGGNet-19, ResNet50, and Inception_ResNet_V2. Experiment results show that Inception_ResNet_V2 gives better results than VGGNet and ResNet models with training and test accuracy of 99.26 and 94, respectively. [ABSTRACT FROM AUTHOR]

In order to alleviate the increasingly serious traffic congestion problem in China, realize intelligent traffic control, and provide accurate and real-time traffic flow prediction data for traffic flow guidance and traffic travel, this paper designs a GPS-based vehicle trajectory fusion optimization deep model BN-LSTM-CNN which makes full use of the temporal and spatial correlation characteristics of dynamic traffic flow to improve the accuracy of short-term traffic flow prediction. The parameters of the historical GPS dynamic trajectory of the traffic network link are converted into a two-dimensional matrix image of time and space relationship. First, the spatial features are input to the CNN network, and the spatial dependence relationship between the links is mined, then the traffic flow time series modeling is performed with a four-layer ConvLSTM network, and the BN normalization layer is added to normalize the activation value of the previous layer on each batch, so that the model can obtain higher training accuracy and quickly complete the prediction of the traffic flow state in a certain period of time in the future. The experimental results show that the prediction model is fast to optimize, the prediction error is the smallest compared with other methods, and it can meet the real-time requirements of urban traffic control. [ABSTRACT FROM AUTHOR]

The prediction of disaster risk paths can foresee the spread of disaster risk and lay a foundation for reducing or avoiding the harm of disaster risk. In this paper, we have improved the embedding layer of convolution neural network (CNN), and we literally refer to the specific CNN framework as IECNN. The IECNN model is proposed to predict the disaster risk path. Here, we first establish a disaster risk network in which each node represents two attributes: place and disaster type. The random walk paths are generated from the disaster risk network, and the structural characteristics of nodes in the disaster risk paths are considered in the improved embedding layer of IECNN. This study also applies Markov chain algorithm to calculate the probability of each disaster risk path and trains the IECNN model with the determined disaster risk paths. ROC curve, AUC, Accuracy, F1-measure, Precision, and Recall are used to evaluate the prediction model. In order to verify the feasibility and advantage of the proposed model, we use a dataset consisting of natural disasters in southwest China in the experimental section. Results of the comparative analysis display that the proposed model cannot only effectively predict the disaster risk path, but also provide the high performance in terms of every evaluation index. Moreover, some characters of disaster risk path are found and discussed. Resultantly, our results can provide an efficient way to prevent and control risk spread in disaster events. [ABSTRACT FROM AUTHOR]

Landslides are one of the main geological disasters in the riparian zone of the Three Gorges Reservoir Area (TGRA) in China, causing massive losses in terms of industrial and agricultural production and people's lives and property. Due to the difficulty of data collection and lack of facade information, the remote sensing methods currently used for landslide mapping and detection are insufficient. The current landslide information acquisition method is mainly based on visual interpretation and automatic computer classification from aerial orthophotographs, which is subjective and time-consuming and yields low accuracy. New shipborne photogrammetry has unique advantages in small and steep slope landslide detection. In recent years, the deep learning technology represented by convolutional neural networks (CNNs) exhibits more powerful feature learning ability in image recognition than traditional machine learning technology. Meanwhile, transfer learning can make the network suitable for small datasets through fine-tuning, which could transfer the parameters of the pretrained model to the new model in a certain way, thereby increasing the learning efficiency of the model. In this paper, the training of multiple deep learning models is realized through transfer learning, and automatic identification of landslides based on shipborne images is achieved through decision-level fusion. The experimental results show that through the integration of decision-level fusion and transfer learning, it performs well in the classification of landslides based on shipborne images. [ABSTRACT FROM AUTHOR]

The high-speed train (HST) is one of the most important transport tools in China, and the sound quality of its interior noise affects passengers' comfort. This paper proposes a HST sound quality model. The model combines Mel-scale frequency cepstral coefficients (MFCCs), the most popular spectral-based input parameter in deep learning models, with convolutional neural networks (CNNs) to evaluate the sound quality of HSTs. Meanwhile, two input channels are applied to simulate binaural hearing so that the different sound signals can be processed separately. The binaural MFCC-CNN model achieves an accuracy of 96.2% and outperforms the traditional shallow neural network model because it considers the time-varying characteristics of noise. The MFCC features are capable of capturing the characteristics of noise and improving the accuracy of sound quality evaluations. Besides, the results suggest that the time and level differences in sound signals are important factors affecting sound quality at low annoyance levels. The proposed model is expected to optimize the comfort of the interior acoustic environment of HSTs. [ABSTRACT FROM AUTHOR]

Accurately predicting wind and photovoltaic power is one of the keys to improving the economy of wind-solar complementary power generation system, reducing scheduling costs and no-load losses, and ensuring grid stability. However, the natural properties of energy result in complex fluctuations in their corresponding power sequences, making accurate predictions difficult. Therefore, this paper proposes an intelligent prediction system that combines decomposition algorithms and deep learning for ultra-short-term prediction of wind and photovoltaic power. First, an improved decomposition algorithm is proposed, based on fuzzy entropy's property that its value increases with the increase of sequence uncertainty, particle swarm optimization (PSO) is used to search for the optimal parameter combinations of variational modal decomposition (VMD), so that it can automatically adjust the parameters for energy data with different characteristics to reduce the human error. Then, a convolutional neural network (CNN) architecture that balances operational efficiency and prediction performance is constructed, and the hyperparameters of the CNN are optimized using the wild horse optimization algorithm (WHO) to improve the stability and accuracy of the prediction model. In this paper, real data from wind power plants and photovoltaic power plants in China are used as experimental objects, and experiments are carried out in three aspects, namely, benchmark model selection, decomposition algorithm comparison and combined model comparison. The results show that selecting CNN as the benchmark model is a good choice; the improved VMD has better decomposition performance than other state-of-the-art decomposition algorithms. The system proposed in this paper is highly generalizable and adaptive, and its prediction performance and accuracy greatly outperform that of other comparative models, with prediction accuracies improved by 72% and 79%, respectively, compared to a single CNN model. • The hyperparameters of VMD are determined by using PSO based on fuzzy entropy. • Optimize convolutional neural network using the wild horse optimization algorithm. • The intelligent prediction system can accurately predict wind power and PV power. • Experiments based on power data from actual wind farms and PV plants. • A deep learning prediction method applied to wind and solar complementary systems. [ABSTRACT FROM AUTHOR]

This study develops a model to assess building vulnerability across Xinxing County by integrating quantitative derivation with machine learning techniques. Building vulnerability is characterized as a function of landslide hazard risk and building resistance, wherein landslide hazard risk is derived using CNN (1D) for nine hazard-causing factors (elevation, slope, slope shape, geotechnical body type, geological structure, vegetation cover, watershed, and land-use type) and landslide sites; building resistance is determined through quantitative derivation. After evaluating the building susceptibility of all the structures, the susceptibility of each village is then calculated through subvillage statistics, which are aimed at identifying the specific needs of each area. Simultaneously, different landslide hazard classes are categorized, and an analysis of the correlation between building resistance and susceptibility reveals that building susceptibility exhibits a positive correlation with landslide hazard and a negative correlation with building resistance. Following a comprehensive assessment of building susceptibility in Xinxing County, a sample encompassing different landslide intensity areas and susceptibility classes of buildings was chosen for on-site validation, thus yielding an accuracy rate of the results as high as 94.5%. [ABSTRACT FROM AUTHOR]

Since the early 1950s, the development of human settlements and over-exploitation of agriculture in the China side of the Amur River Basin (CARB) have had a major impact on the water environment of the surrounding lakes, resulting in a decrease of aquatic vegetation. According to the United Nations Sustainable Development Goals, a comprehensive understanding of the extent and variability of aquatic vegetation is crucial for preserving the structure and functionality of stable aquatic ecosystems. Currently, there is a deficiency in the CARB long-sequence dataset of aquatic vegetation distribution in China. This shortage hampers effective support for actual management. Therefore, the development of a fast, robust, and automatic method for accurate extraction of aquatic vegetation becomes crucial for large-scale applications. Our objective is to gather information on the spatial and temporal distribution as well as changes in aquatic vegetation within the CARB. Utilizing a hybrid approach that combines the maximum spectral index composite and Otsu algorithm, along with the integration of convolutional neural networks (CNN) and random forest, we applied this methodology to obtain an annual dataset of aquatic vegetation spanning from 1985 to 2020 using Landsat series imagery. The accuracy of this method was validated through both field investigations and Google Images. Upon assessing the confusion matrix spanning from 1985 to 2020, the producer accuracy for aquatic vegetation classification consistently exceeded 87%. Further quantitative analysis unveiled a discernible decreasing trend in both the water and vegetation areas of lakes larger than 20 km2 within the CARB over the past 36 years. Specifically, the total water area decreased from 3575 km2 to 3412 km2, while the vegetation area decreased from 745 km2 to 687 km2. These changes may be attributed to a combination of climate change and human activities. These quantitative data hold significant practical implications for establishing a scientific restoration path for lake aquatic vegetation. They are particularly valuable for constructing the historical background and reference indices of aquatic vegetation. [ABSTRACT FROM AUTHOR]

Drought is a natural disaster that occurs globally and can damage the environment, disrupt agricultural production and cause large economic losses. The accurate prediction of drought can effectively reduce the impacts of droughts. Deep learning methods have shown promise in drought prediction, with convolutional neural networks (CNNs) being particularly effective in handling spatial information. In this study, we employed a deep learning approach to predict drought in the Fenhe River (FHR) basin, taking into account the meteorological conditions of surrounding regions. We used the daily SAPEI (Standardized Antecedent Precipitation Evapotranspiration Index) as the drought evaluation index. Our results demonstrate the effectiveness of the CNN model in predicting drought events 1~10 days in advance. We evaluated the predictions made by the model; the average Nash–Sutcliffe efficiency (NSE) between the predicted and true values for the next 10 days was 0.71. While the prediction accuracy slightly decreased with longer prediction lengths, the model remained stable and effective in predicting heavy drought events that are typically difficult to predict. Additionally, key meteorological variables for drought predictions were identified, and we found that training the CNN model with these key variables led to higher prediction accuracy than training it with all variables. This study approves an effective deep learning approach for daily drought prediction, particularly when considering the meteorological conditions of surrounding regions. [ABSTRACT FROM AUTHOR]

The quantity of photovoltaic systems generated relies on the climate, and to generate significant variations depending on the climatic conditions. Its electricity production processes exhibit unpredictability, instability, and intermittent behavior. The augmented convolutional long-short memory combination network is recommended for PV energy combined forecasting in accordance to overcome the challenge of considerable instability in the high proportion of sustainable energy output. The LSTM and CNN models' training procedure uses the enhanced artificial bee colony (ABC) to adjust their characteristics. To determine the projected model parameters of PV power, the estimated input parameters of every model are stacked and then reassembled. The studies showed that the forecast precision was significantly higher than the reliability rate of the Convolutional LSTM combinational network as well as the forecasting findings from the BP, CNN, and LSTM independent computational methods. This forecast specificity was predicated on the enhanced ABC to optimize the variables of the CLSTM combinational network for forecasting Photo Voltaic power. For comparative study and assessment, the most recent operational data from a Photovoltaic power plant in northern China have been utilized. The suggested technique resolved more quickly and was higher weather-adaptable than the individual neural network and hybrid forecasting models mentioned prior. [ABSTRACT FROM AUTHOR]

Severe Acute Respiratory Syndrome Corona Virus 2 or SARS-CoV-2 (COVID-19) has affected 21 million people worldwide and is responsible for 0.75 million deaths (as of August 2020). Declared as a pandemic by the WHO, the virus has affected almost every country after originating in China articulating how contagious the virus is. With reasonable social measures, countries have already shown depreciation in coronavirus cases. This paper deals with detecting and distinguishing the COVID-19 disease from normal patients through frontal chest X-ray scans using Convolutional Neural Networks. Different combinations of parameters were used to train multiple CNNs and their behavior was noted. After thorough experimentation of different models, the best model which achieved an accuracy of 0.98 on the test set with a loss of 0.036 was nominated. The selected model's more intuitive accuracy metrics were calculated and its intermediate convolutional neural activations and receiver-operating characteristics curve is shown. [ABSTRACT FROM AUTHOR]

Objective: Large-scale human engineering activities induce and aggravate the landslide disaster, which seriously threatens the engineering safety and environmental safety. Landslide susceptibility assessment is the key technique of landslide monitoring and early warning. In view of the limited data sources of traditional landslide monitoring Methods: and the lack of effective Methods: for mining the spatial distribution of landslide hazards and their inducing factors, the main Objective: of this study was to investigate the prediction performances of synthetic minority oversampling technique (SMOTE) and convolutional neural networks (CNN) for landslide susceptibility assessment in the Wanzhou area of the Three Gorges, China. Methods:: The landslide susceptibility assessment Methods: used in this paper are SMOTE and CNN. SMOTE is an improved random oversampling algorithm, which was used to better the imbalance of input samples. CNN is a widely used deep learning network and differs from other neural networks mainly in convolution operation. CNN was used to analyze the nonlinear relationship between landslides and its influencing factors. Firstly, 22 influencing factors were extracted from multi-source data such as topographic, geological, and remote sensing data, and multicollinearity test was performed on these factors. Then, the grid unit with a spatial resolution of 25 m was selected as the mapping unit. 30% of the initial sample dataset was selected randomly for testing of the CNN, and the remaining 70% was expanded by using SMOTE to obtain 1:1 landslide and non-landslide dataset. 90% of the proportionally balanced dataset was randomly selected as the training set and the remaining 10% as the validation set. Thirdly, the obtained training sam - ple set is input to the CNN and six optimal model parameters were set to ensure the stability and efficiency of the model. Finally, the trained CNN was used to generate landslide susceptibility zoning map, and the receiver operating characteristic curves, testing and validating set were used for evaluating the accuracy and Results of the proposed model.Results The proposed SMOTE-CNN model with the highest predictive accuracy and generalization ability was trained and then used to calculate the landslide susceptibility indices. The index values of the grid cells vary from zero to one and correspond to landslide susceptibilities from low to high. The landslide susceptibility maps were created using these values. For easy visual interpretation, the susceptibility values were classified into five classes (very low,low,moderate,high,and very high) using Jenks natural breaks. Approximately, 59.57 % of the landslides lie in the very-high-susceptibility region, 18.74% of the landslides lie in the high-susceptibility region, 10.73% of the landslides lie in the moderatesusceptibility region, and 10.96 % of the landslides lie in the low- and very-low-susceptibility regions. The experimental Results demonstrate that the proposed model provides the best predictive accuracy. The model can effectively assess landslide susceptibility and provides a novel method for landslide prediction.Conclusions: The SMOTE and CNN were applied to quantitatively mapping the landslide susceptibility in Wanzhou area of the Three Gorges, China. The experimental Results: show that SMOTE and CNN have good stability, and the Results were consistent with field investigations and can provide a reference for landslide prevention and reduction in the Three Gorge, China. In addition, CNN has the characteristics of strong adaptability and is good at mining local features of data and extracting global training features and classification. However, the physical meaning of CNN model is not clear, and it takes a long time to process a large amount of data, so it relies on GPU acceleration. Therefore, it is necessary to carry out further research from the aspects of super-parameter optimization combination, model structure adjustment, applicability analysis etc., so as to improve the accuracy and robustness of the model. [ABSTRACT FROM AUTHOR]

In many regions of the world, early diagnosis of non-small cell lung cancer (NSCLC) is a major challenge due to the large population and lack of medical resources, which is difficult toeffectively address via limited physician manpower alone. Therefore, we developed a convolutional neural network (CNN)-based assisted diagnosis and decision-making intelligent medical system with sensors. This system analyzes NSCLC patients' medical records using sensors to assist staging a diagnosis and provides recommended treatment plans to physicians. To address the problem of unbalanced case samples across pathological stages, we used transfer learning and dynamic sampling techniques to reconstruct and iteratively train the model to improve the accuracy of the prediction system. In this paper, all data for training and testing the system were obtained from the medical records of 2,789,675 patients with NSCLC, which were recorded in three hospitals in China over a five-year period. When the number of case samples reached 8000, the system achieved an accuracy rate of 0.84, which is already close to that of the doctors (accuracy: 0.86). The experimental results proved that the system can quickly and accurately analyze patient data and provide decision information support for physicians. [ABSTRACT FROM AUTHOR]

• SL, DL and DRL models were constructed for multi performance prediction of gymnasiums under different weather conditions. • Accuracy, time, interpretability, ease-of-use, hyperparameters, and robustness of the metamodels have been investigated. • Energy consumption, temperature and CO 2 time series vectors prediction were modeling with nine training sets. • Methodology for hyperparameters comparison using exhaustive grid search and sensitivity analysis. • RDPG is most accurate but time-consuming. LSTM is more efficient and also has relatively excellent accuracy. Building performance simulation (BPS) is essential for testing energy demand and indoor environment quality of different building designs. However, software for BPS is computationally intensive and impractical to run thousands even millions of simulations for performance analysis and optimization. Especially for the large space buildings, which usually have complex forms and high energy consumption. The computational problem could be overcome by the adoption of metamodels. Most correlational research focuses on single performance prediction of a particular building, which makes the model less robust when applied to multi vectors prediction for different buildings under multiple weather conditions. This paper leveraged six metamodels, which include recurrent deterministic policy gradient (RDPG), asynchronous advantage actor-critic (A3C), long short-term memory (LSTM), convolution neural network (CNN), artificial neural network (ANN) and support vector regression (SVR), to predict hourly-based multi performance vectors of gymnasiums under various design parameters and multiple weather conditions. Six metamodels are trained and tested on a large scale of datasets simulated by EnergyPlus over four gymnasium cases in different cities of China. The accuracy, efficiency, ease-of-use, robustness and interpretability of the models are investigated. To conduct a fair and detailed comparison, a methodological approach using grid searches for model settings selection assisted by sensitivity analysis is pursued. Principal component analysis (PCA) is also adopted to interpret the work process of the metamodels. The comparison showed that the RDPG model provides the most accurate prediction results with R2 converges at 0.993, 0.982 and 0.941 for energy, temperature and CO 2 , respectively. LSTM model is more efficient than RDPG, and suitable for users who need emphasis on both time and accuracy. ANN is suitable for users with limited time and require models of ease-of-use and robustness. SVR and ANN could be used for the automatically co-simulation with BPS software. In future research, the influence of occupant behavior also should be investigated. [ABSTRACT FROM AUTHOR]

The construction of smart grid is inseparable from the appropriate deployment of utility poles. China has many poles; if they are fully utilized, can they bring more power to the smart grid? Fortunately, thanks to advances in technologies such as artificial intelligence and the Internet of Things, can we use Google Earth to recognize utility poles and optimize the deployment of power facilities? Fundamentally, Google Earth generates static images. Faced with many static images, we need further image processing for high-definition recognition of utility poles. In addition to the current chaotic distribution of poles, we also need to optimize the deployment of power facilities. However, due to the resolution of Google Earth and the aerial photography angle, many backlight phenomena are not conducive to the recognition of utility poles. Therefore, this paper proposes a backlight image enhancement algorithm based on convolutional neural networks (CNN) and constructs a novel network architecture that integrates decomposition, restoration, and adjustment to recognize poles in high-definition under backlight. Furthermore, to solve the problems of the overflow of CNN parameters and unclear training effect, particle swarm optimization (PSO), the evolutionary computing-based machine learning (ECML) is used to search CNN parameters automatically and seek the optimal solution to achieve the optimization of the overall model. The experiments prove that the CNN-based image enhancement algorithm effectively recognizes the utility poles under different illumination and backlight. At the same time, the experimental results show that the PSO-based optimization method can optimize CNN parameters obviously, and the classification accuracy is increased. • An image enhancement algorithm based on CNN for Google Earth static images is proposed for recognition of utility poles. • A novel network architecture integrating decomposition, restoration, and adjustment is constructed. • A PSO-based method is proposed to automatically tune the parameters of CNN model. [ABSTRACT FROM AUTHOR]

The novel coronavirus infection (COVID-19) that was first identified in China in December 2019 has spread across the globe rapidly infecting over ten million people. The World Health Organization (WHO) declared it as a pandemic on March 11, 2020. What makes it even more critical is the lack of vaccines available to control the disease, although many pharmaceutical companies and research institutions all over the world are working toward developing effective solutions to battle this life-threatening disease. X-ray and computed tomography (CT) images scanning is one of the most encouraging exploration zones; it can help in finding and providing early diagnosis to diseases and gives both quick and precise outcomes. In this study, convolution neural networks method is used for binary classification pneumonia-based conversion of VGG-19, Inception_V2 and decision tree model on X-ray and CT scan images dataset, which contains 360 images. It can infer that fine-tuned version VGG-19, Inception_V2 and decision tree model show highly satisfactory performance with a rate of increase in training and validation accuracy (91%) other than Inception_V2 (78%) and decision tree (60%) models. [ABSTRACT FROM AUTHOR]

Corn is one of the main food crops in China, and its area ranks in the top three in the world. However, the corn leaf disease has seriously affected the yield and quality of corn. To quickly and accurately identify corn leaf diseases, taking timely and effective treatment to reduce the loss of corn yield. We proposed identifying corn leaf diseases using the Mobilenetv3 (CD-Mobilenetv3) model. Based on the Mobilenetv3 model, we replaced the model's cross-entropy loss function with a bias loss function to improve accuracy. Replaced the model's squeeze and excitation (SE) module with the efficient channel attention (ECA) module to reduce parameters. Introduced the cross-layer connections between Mobile modules to utilize features synthetically. Then we Introduced the dilated convolutions in the model to increase the receptive field. We integrated a hybrid open-source corn leaf disease dataset (CLDD). The test results on CLDD showed the accuracy reached 98.23%, the precision reached 98.26%, the recall reached 98.26%, and the F1 score reached 98.26%. The test results are improved compared to the classic deep learning (DL) models ResNet50, ResNet101, ShuffleNet_x2, VGG16, SqueezeNet, InceptionNetv3, etc. The loss value was 0.0285, and the parameters were lower than most contrasting models. The experimental results verified the validity of the CD-Mobilenetv3 model in the identification of corn leaf diseases. It provides adequate technical support for the timely control of corn leaf diseases. [ABSTRACT FROM AUTHOR]

In the past, reservoir engineers used numerical simulation or reservoir engineering methods to predict oil production, and the accuracy of prediction depended more on the engineers' own experience. With the development of data science, a new trend has arisen to use deep learning to predict oil production from the perspective of data. In this study, a hybrid forecasting model (CNN-LSTM) based on a convolutional neural network (CNN) and a Long Short-Term Memory (LSTM) neural network is proposed and used to predict the production of fractured horizontal wells in volcanic reservoirs. The model solves the limitation of traditional methods that rely on personal experience. First, the production constraints and production data are used to form a feature space, and the abstract semantics of the feature time series are extracted through convolutional neural network, then the LSTM neural network is used to predict the time series. The certain hyperparameters of the whole model are optimized by Particle Swarm Optimization algorithm (PSO). In order to estimate the model, some production dynamics from the Xinjiang oilfield of China are used for comparative analysis. The experimental results show that the CNN-LSTM model is superior to traditional neural networks and conventional decline curves. [ABSTRACT FROM AUTHOR]

The urban agglomeration area is a heavy disaster area of PM2.5 pollution, and the problem of PM2.5 pollution seriously affects the natural environment and public health. Accurate prediction of PM2.5 concentrations in urban agglomerations is the basis for effective environmental management. Deep learning methods such as Bi-directional Long Short-Term Memory (BiLSTM) and Convolutional Neural Network (CNN) have been applied to PM2.5 concentration prediction in urban agglomerations. However, the existing research methods have low prediction accuracy and poor practicality, and lack a universal method to accurately predict PM2.5 concentrations in 19 urban agglomerations of China. Based on this, this paper combines BiLSTM and CNN, and introduces a Learning Rate Schedule (LRS) to propose a hybrid model of LRS + BiLSTM + CNN. This model combines the advantages of BiLSTM and CNN to extract temporal and spatial features of PM2.5 data. In addition, LRS is introduced in the process of model training, which dynamically adjusts the learning rate of the network model and reduces the training cost of the model. Finally, the daily average datasets of six air qualities from nineteen urban agglomerations of China are used as examples, and seven benchmark network models, CNN, Long Short-Term Memory, BiLSTM, CNN + LSTM, LSTM + CNN, and CNN + BiLSTM, are considered for validation experiments. The results show that the prediction effect of the proposed model is significantly better than that of the benchmark model in 17 urban agglomerations, such as Mid-southern Liaoning urban agglomerations and Central Yunnan urban agglomeration. Compared with the above seven benchmark models, the prediction accuracy is improved by 12.7%, 8.9%, 12.4%, 12.8%, 16.4%, 19.7%, 9.5%; 10.2%, 24.2%, 14.4%, 17.9%, 18.4%, 19.9% and 2.6% respectively. The LRS is fully applicable to nineteen urban agglomerations, which has clear effects on the optimization of the network model. LRS + BiLSTM + CNN has high accuracy and practicality in predicting PM2.5 concentrations in urban agglomeration areas in China, and can provide effective technical support for environmental governance of urban agglomeration areas. [ABSTRACT FROM AUTHOR]

The precise brain tumor diagnosis is critical and shows a vital role in the medical support for treating tumor patients. Manual brain tumor segmentation for cancer analysis from many Magnetic Resonance Images (MRIs) created in medical practice is a problematic and timewasting task for experts. As a result, there is a critical necessity for more accurate computeraided methods for early tumor detection. To remove this gap, we enhanced the computational power of a computer-aided system by proposing a finetuned Block-Wise Visual Geometry Group19 (BW-VGG19) architecture. In this method, a pre-trained VGG19 is fine-tuned with CNN architecture in the block-wise mechanism to enhance the system's accuracy. The publicly accessible Contrast-Enhanced Magnetic Resonance Imaging (CE-MRI) dataset collected from 2005 to 2020 from different hospitals in China has been used in this research. Our proposed method is simple and achieved an accuracy of 0.98%. We compare our technique results with the existing Convolutional Neural network (CNN), VGG16, and VGG19 approaches. The results indicate that our proposed technique outperforms the best results associated with the existing methods. [ABSTRACT FROM AUTHOR]

Open caissons are widely used in civil engineering as deep foundations and underground structures. It is necessary to control the sinking process of open caissons to achieve safe sinking. This paper proposes a key index of open caisson sinking, the sinking by earth excavation index (SEI). Then, a time series prediction model of SEI based on convolutional neural network (CNN) was established, and an active control method with a data-driven strategy was proposed based on this model. The proposed model was applied in the Changtai Yangtze River Bridge Project in China and was validated by field data. Finally, the effects of hyperparameters in the model were analysed, and a real-time SEI prediction was simulated to describe the practicability of the model. Results show that the proposed model is reliable and practicable in open caisson engineering, and the prediction accuracy is higher than that of other algorithms. Three hyperparameters have little effect on prediction accuracy, including the number of neurons in the fully connected layer, dropping probability of dropout and input length, while the prediction length has a strong influence. With strong objectivity and remarkable performance, the proposed method is conducive to ensuring the steady state and safety of open caissons. • A key index is proposed for the sinking control of open caissons. • A CNN-based intelligent model of the key index is established. • The model is validated to be applicable in the open caisson engineering. • The active control method is proposed based on the model. [ABSTRACT FROM AUTHOR]

In this study, a deep learning model is proposed to predict groundwater levels. The model is able to accurately complete the prediction task even when the data utilized are insufficient. The hybrid model that we have developed, CNN-LSTM-ML, uses a combined network structure of convolutional neural networks (CNN) and long short-term memory (LSTM) network to extract the time dependence of groundwater level on meteorological factors, and uses a meta-learning algorithm framework to ensure the network's performance under sample conditions. The study predicts groundwater levels from 66 observation wells in the middle and lower reaches of the Heihe River in arid regions and compares them with other data-driven models. Experiments show that the CNN-LSTM-ML model outperforms other models in terms of prediction accuracy in both the short term (1 month) and long term (12 months). Under the condition that the training data are reduced by 50%, the MAE of the proposed model is 33.6% lower than that of LSTM. The results of ablation experiments show that CNN-LSTM-ML is 26.5% better than the RMSE of the original CNN-LSTM structure. The model provides an effective method for groundwater level prediction and contributes to the sustainable management of water resources in arid regions. [ABSTRACT FROM AUTHOR]

Mu Us Sandy Land is a typical semi-arid vulnerable ecological zone, characterized by vegetation degradation and severe desertification. Effectively identifying desertification changes has been a topical environmental issue in China. However, most previous studies have used a single method or remote sensing index to monitor desertification, and lacked an efficient and high-precision monitoring system. In this study, an optimal monitoring scheme that considers multiple indicators combination and different machine learning methods (Classification and Regression Tree-Decision Tree, CART-DT; Random Forest, RF; Convolutional Neural Networks, CNN) was developed and used to analyze the spatial–temporal patterns of desertification from 2000 to 2018 in Mu Us Sandy Land. The results showed that: (a) The random forest model performed best for monitoring desertification based on medium and low-resolution remote sensing images, and the four-index combination (Albedo, NDVI, LST and TGSI) obtained the highest classification accuracy (OA = 87.67%) in Mu Us Sandy Land. Surprisingly, the model accuracy of the three-index combination (NDVI, LST and TGSI) (OA = 85.74%) is comparable to the four-index combination. (b) The TGSI index used to characterize soil information performs well, while the LST is not conducive to the extraction of desertified land in several desertification monitoring indicators. (c) Since 2000, the area of extremely severe desertified land has shown a reversal trend; however, there is significant interannual fluctuation in the total and light desertification land area affected by extreme climate. This research provides a novel approach and a valuable reference for monitoring the evolution of desertification in regional studies, and the results improve the research system of desertification and provide a data basis for desertification cause analysis and prevention. [ABSTRACT FROM AUTHOR]

It is difficult to improve the seasonal prediction skill of winter temperature over North China, owing to the complex dynamics of East Asian winter and the relatively low prediction skill level of current climate models. Deep learning (DL) may be an informative and promising tool to enhance seasonal prediction, particularly in regions where the underlying mechanisms are not clear. Here, using a DL model based on the Convolutional Neural Network (CNN), we have found that the prediction skill for North China winter temperature (NCWT) can be extended up to five months by considering the remote impact of the Northeast Pacific sea-surface temperature (SST) on North China. Based on historical simulations of winter temperatures in North China, we selected six CMIP5 models with relatively small deviations for training the CNN, and the period chosen for training was 1852–1991. The ERA5 data during 1995–2017 were utilized to evaluate the performance of the CNN. Our CNN shows the best performance in a recent 10-year period (2008–2017), showing a significantly improved level of NCWT prediction skill with a correlation skill of 0.65 at a 5-month lead time, which is much better than the forecast skill of the state-of-the-art dynamic seasonal prediction system. Heat map analysis was used to explore the possible physical mechanisms associated with the NCWT anomaly from the perspective of the CNN; the results showed that the SST over the Northeast Pacific is highly relevant to NCWT prediction. The Northeast Pacific warming in the boreal summer is related to the development of the El Niño event in the coming winter, which may induce NCWT anomalies by atmospheric teleconnection. Climate model experiments support the role of Northeast Pacific warming in the boreal summer on NCWT. The improved capability for prediction from using the CNN may help to establish the energy policy for the coming winter and reduce the economic losses from extremely cold in North China. [ABSTRACT FROM AUTHOR]

Online ride-hailing order forecasting is a very important part of the intelligent traffic dispatch system. Accurate order forecasting can reduce the flow of invalid vehicles and improve the user experience of online ride-hailing. We propose a multi-view deep long short-term memory (LSTM) network architecture (MultiView deep LSTM framework), which uses convolutional neural network and graph convolutional network to extract the temporal and spatial characteristics of online ride-hailing orders, obtains the correlation information between regional orders through the order view, regional speed view, and weather factor view, and then uses LSTM unit and attention unit to predict the order volume in real time. We use Didi Haikou, China's online ride-hailing dataset for training, compare it with the prediction algorithms of other articles, and experiment with different choices of the contrast framework. The experimental results show that our deep learning framework can effectively capture comprehensive spatio-temporal correlation and obtain better results. The model maintained good performance at 15 min, 30 min, and 1 h. Experiments conducted on the actual demand data onto ride-hailing from Didi Haikou data prove that our method is better than the latest method. [ABSTRACT FROM AUTHOR]

The increasing volume of seismic data from long-term continuous monitoring motivates the development of algorithms based on convolutional neural network (CNN) for faster and more reliable phase detection and picking. However, many less studied regions lack a significant amount of labeled events needed for traditional CNN approaches. In this paper, we present a CNN-based Phase-Identification Classifier (CPIC) designed for phase detection and picking on small to medium sized training datasets. When trained on 30,146 labeled phases and applied to one-month of continuous recordings during the aftershock sequences of the 2008 M W 7.9 Wenchuan Earthquake in Sichuan, China, CPIC detects 97.5% of the manually picked phases in the standard catalog and predicts their arrival times with a five-times improvement over the ObsPy AR picker. In addition, unlike other CNN-based approaches that require millions of training samples, when the off-line training set size of CPIC is reduced to only a few thousand training samples the accuracy stays above 95%. The deployment of CPIC takes less than 12 h to pick arrivals in 31-day recordings on 14 stations. In addition to the catalog phases manually picked by analysts, CPIC finds more phases for existing events and new events missed in the catalog. Among those additional detections, some are confirmed by a matched filter method while others require further investigation. Finally, when tested on a small dataset from a different region (Oklahoma, US), CPIC achieves 97% accuracy after fine tuning only the fully connected layer of the model. This result suggests that the CPIC developed in this study can be used to identify and pick P/S arrivals in other regions with no or minimum labeled phases. • Design and test a feasible CNN scheme for small to medium size training sets which is accurate and efficient compared to manual picking. • The aftershock dataset used in the test is known to be challenging for phase picking due to the existence of a large number of weak events and low signal-to-noise ratio. • The CNN structure we propose can be trained on the small training set (< 10, 000 labels) with 95% classiffication accuracy. • Deployment of the trained CNN is efficient: detection takes two hours and picking takes 12 hours to process 31 days of data on 14 stations. • Many weak events detected by the CNN but missed by manual pickings can be verified by a more complete catalog created via the matched filter approach. [ABSTRACT FROM AUTHOR]

Image classification aims to automatically group a set of images into several categorizations, which is widely applied in scene categorization, image clustering. Lung cancer recognition can be achieved by using image classification technique, since there are distinct differences between healthy lung and sick lung images. In this paper, we propose lung cancer recognition based on image quality assessment, which can distinguish sick lung images from healthy lung images. First, our dataset is acquired using low-dose CT scan combined with full-mode iterative recombination (IMR). Then, we incorporate both low-level and high-level features to extract deep representation from obtained dataset. Specifically, our designed low-level features include color moment and texture feature, and CNN based method is leveraged for deep feature extraction. For reducing artifacts and noise of images, we assign quality score for each training image. And quality score and deep feature are fused to generate deep representation. Afterward, we propose a probabilistic model to learn the distribution of deep representation. Finally, lung cancer recognition can be achieved using learned model. We conduct comprehensive experiments and our proposed method is verified effective. [ABSTRACT FROM AUTHOR]

Deep learning techniques have been successfully applied in handwriting recognition. Oracle bone inscriptions (OBI) are the earliest hieroglyphs in China and valuable resources for studying the etymology of Chinese characters. OBI are of important historical and cultural value in China; thus, textual research surrounding the characters of OBI is a huge challenge for archaeologists. In this work, we built a dataset named OBI-100, which contains 100 classes of oracle bone inscriptions collected from two OBI dictionaries. The dataset includes more than 128,000 character samples related to the natural environment, humans, animals, plants, etc. In addition, we propose improved models based on three typical deep convolutional network structures to recognize the OBI-100 dataset. By modifying the parameters, adjusting the network structures, and adopting optimization strategies, we demonstrate experimentally that these models perform fairly well in OBI recognition. For the 100-category OBI classification task, the optimal model achieves an accuracy of 99.5%, which shows competitive performance compared with other state-of-the-art approaches. We hope that this work can provide a valuable tool for character recognition of OBI. [ABSTRACT FROM AUTHOR]

Accurate crop type maps play an important role in food security due to their widespread applicability. Optical time series data (TSD) have proven to be significant for crop type mapping. However, filling in missing information due to clouds in optical imagery is always needed, which will increase the workload and the risk of error transmission, especially for imagery with high spatial resolution. The development of optical imagery with high temporal and spatial resolution and the emergence of deep learning algorithms provide solutions to this problem. Although the one-dimensional convolutional neural network (1D CNN), long short-term memory (LSTM), and gate recurrent unit (GRU) models have been used to classify crop types in previous studies, their ability to identify crop types using optical TSD with missing information needs to be further explored due to their different mechanisms for handling invalid values in TSD. In this research, we designed two groups of experiments to explore the performances and characteristics of the 1D CNN, LSTM, GRU, LSTM-CNN, and GRU-CNN models for crop type mapping using unfilled Sentinel-2 (Sentinel-2) TSD and to discover the differences between unfilled and filled Sentinel-2 TSD based on the same algorithm. A case study was conducted in Hengshui City, China, of which 70.3% is farmland. The results showed that the 1D CNN, LSTM-CNN, and GRU-CNN models achieved acceptable classification accuracies (above 85%) using unfilled TSD, even though the total missing rate of the sample values was 43.5%; these accuracies were higher and more stable than those obtained using filled TSD. Furthermore, the models recalled more samples on crop types with small parcels when using unfilled TSD. Although LSTM and GRU models did not attain accuracies as high as the other three models using unfilled TSD, their results were almost close to those with filled TSD. This research showed that crop types could be identified by deep learning features in Sentinel-2 dense time series images with missing information due to clouds or cloud shadows randomly, which avoided spending a lot of time on missing information reconstruction. [ABSTRACT FROM AUTHOR]

Grape is an economic crop of great importance and is widely cultivated in China. With the development of remote sensing, abundant data sources strongly guarantee that researchers can identify crop types and map their spatial distributions. However, to date, only a few studies have been conducted to identify vineyards using satellite image data. In this study, a vineyard is identified using satellite images, and a new approach is proposed that integrates the continuous wavelet transform (CWT) and a convolutional neural network (CNN). Specifically, the original time series of the normalized difference vegetation index (NDVI), enhanced vegetation index (EVI), and green chlorophyll vegetation index (GCVI) are reconstructed by applying an iterated Savitzky-Golay (S-G) method to form a daily time series for a full year; then, the CWT is applied to three reconstructed time series to generate corresponding scalograms; and finally, CNN technology is used to identify vineyards based on the stacked scalograms. In addition to our approach, a traditional and common approach that uses a random forest (RF) to identify crop types based on multi-temporal images is selected as the control group. The experimental results demonstrated the following: (i) the proposed approach was comprehensively superior to the RF approach; it improved the overall accuracy by 9.87% (up to 89.66%); (ii) the CWT had a stable and effective influence on the reconstructed time series, and the scalograms fully represented the unique time-related frequency pattern of each of the planting conditions; and (iii) the convolution and max pooling processing of the CNN captured the unique and subtle distribution patterns of the scalograms to distinguish vineyards from other crops. Additionally, the proposed approach is considered as able to be applied to other practical scenarios, such as using time series data to identify crop types, map landcover/land use, and is recommended to be tested in future practical applications. [ABSTRACT FROM AUTHOR]

With the development of ground-based all-sky airglow imager (ASAI) technology, a large amount of airglow image data needs to be processed for studying atmospheric gravity waves. We developed a program to automatically extract gravity wave patterns in the ASAI images. The auto-extraction program includes a classification model based on convolutional neural network (CNN) and an object detection model based on faster region-based convolutional neural network (Faster R-CNN). The classification model selects the images of clear nights from all ASAI raw images. The object detection model locates the region of wave patterns. Then, the wave parameters (horizontal wavelength, period, direction, etc.) can be calculated within the region of the wave patterns. Besides auto-extraction, we applied a wavelength check to remove the interference of wavelike mist near the imager. To validate the auto-extraction program, a case study was conducted on the images captured in 2014 at Linqu (36.2°N, 118.7°E), China. Compared to the result of the manual check, the auto-extraction recognized less (28.9% of manual result) wave-containing images due to the strict threshold, but the result shows the same seasonal variation as the references. The auto-extraction program applies a uniform criterion to avoid the accidental error in manual distinction of gravity waves and offers a reliable method to process large ASAI images for efficiently studying the climatology of atmospheric gravity waves. [ABSTRACT FROM AUTHOR]

A convolutional neural network (CNN) algorithm was developed to retrieve the land surface temperature (LST) from Advanced Microwave Scanning Radiometer 2 (AMSR2) data in China. Reference data were selected using the Moderate Resolution Imaging Spectroradiometer (MODIS) LST product to overcome the problem related to the need for synchronous ground observation data. The AMSR2 brightness temperature (TB) data and MODIS surface temperature data were randomly divided into training and test datasets, and a CNN was constructed to simulate passive microwave radiation transmission to invert the surface temperature. The twelve V/H channel combinations (7.3, 10.65, 18.7, 23.8, 36.5, 89 GHz) resulted in the most stable and accurate CNN retrieval model. Vertical polarizations performed better than horizontal polarizations; however, because CNNs rely heavily on large amounts of data, the combination of vertical and horizontal polarizations performed better than a single polarization. The retrievals in different regions indicated that the CNN accuracy was highest over large bare land areas. A comparison of the retrieval results with ground measurement data from meteorological stations yielded R2 = 0.987, RMSE = 2.69 K, and an average relative error of 2.57 K, which indicated that the accuracy of the CNN LST retrieval algorithm was high and the retrieval results can be applied to long-term LST sequence analysis in China. [ABSTRACT FROM AUTHOR]