Disasters caused by mine water inflows significantly threaten the safety of coal mining operations. Deep mining complicates the acquisition of hydrogeological parameters, the mechanics of water inrush, and the prediction of sudden changes in mine water inflow. Traditional models and singular machine learning approaches often fail to accurately forecast abrupt shifts in mine water inflows. This study introduces a novel coupled decomposition-optimization-deep learning model that integrates Complete Ensemble Empirical Mode Decomposition with Adaptive Noise (CEEMDAN), Northern Goshawk Optimization (NGO), and Long Short-Term Memory (LSTM) networks. We evaluate three types of mine water inflow forecasting methods: a singular time series prediction model, a decomposition-prediction coupled model, and a decomposition-optimization-prediction coupled model, assessing their ability to capture sudden changes in data trends and their prediction accuracy. Results show that the singular prediction model is optimal with a sliding input step of 3 and a maximum of 400 epochs. Compared to the CEEMDAN-LSTM model, the CEEMDAN-NGO-LSTM model demonstrates superior performance in predicting local extreme shifts in mine water inflow volumes. Specifically, the CEEMDAN-NGO-LSTM model achieves scores of 96.578 in MAE, 1.471% in MAPE, 122.143 in RMSE, and 0.958 in NSE, representing average performance improvements of 44.950% and 19.400% over the LSTM model and CEEMDAN-LSTM model, respectively. Additionally, this model provides the most accurate predictions of mine water inflow volumes over the next five days. Therefore, the decomposition-optimization-prediction coupled model presents a novel technical solution for the safety monitoring of smart mines, offering significant theoretical and practical value for ensuring safe mining operations.

Deep learning models provide a more powerful method for accurate and stable prediction of water quality in rivers, which is crucial for the intelligent management and control of the water environment. To increase the accuracy of predicting the water quality parameters and learn more about the impact of complex spatial information based on deep learning models, this study proposes two ensemble models TNX (with temporal attention) and STNX (with spatio-temporal attention) based on seasonal and trend decomposition (STL) method to predict water quality using geo-sensory time series data. Dissolved oxygen, total phosphorus, and ammonia nitrogen were predicted in short-step (1 h, and 2 h) and long-step (12 h, and 24 h) with seven water quality monitoring sites in a river. The ensemble model TNX improved the performance by 2.1%-6.1% and 4.3%-22.0% relative to the best baseline deep learning model for the short-step and long-step water quality prediction, and it can capture the variation pattern of water quality parameters by only predicting the trend component of raw data after STL decomposition. The STNX model, with spatio-temporal attention, obtained 0.5%-2.4% and 2.3%-5.7% higher performance compared to the TNX model for the short-step and long-step water quality prediction, and such improvement was more effective in mitigating the prediction shift patterns of long-step prediction. Moreover, the model interpretation results consistently demonstrated positive relationship patterns across all monitoring sites. However, the significance of seven specific monitoring sites diminished as the distance between the predicted and input monitoring sites increased. This study provides an ensemble modeling approach based on STL decomposition for improving short-step and long-step prediction of river water quality parameter, and understands the impact of complex spatial information on deep learning model.

The primary challenges in tea production under multiple stress exposures have negatively affected its global market sustainability, so introducing an infield fast technique for monitoring tea leaves\' stresses has tremendous urgent needs. Therefore, this study aimed to propose an efficient method for the detection of stress symptoms based on a portable smartphone with deep learning models. Firstly, a database containing over 10,000 images of tea garden canopies in complex natural scenes was developed, which included healthy (no stress) and three types of stress (tea anthracnose (TA), tea blister blight (TB) and sunburn (SB)). Then, YOLOv5m and YOLOv8m algorithms were adapted to discriminate the four types of stress symptoms; where the YOLOv8m algorithm achieved better performance in the identification of healthy leaves (98%), TA (92.0%), TB (68.4%) and SB (75.5%). Furthermore, the YOLOv8m algorithm was used to construct a model for differentiation of disease severity of TA, and a satisfactory result was obtained with the accuracy of mild, moderate, and severe TA infections were 94%, 96%, and 91%, respectively. Besides, we found that CNN kernels of YOLOv8m could efficiently extract the texture characteristics of the images at layer 2, and these characteristics can clearly distinguish different types of stress symptoms. This makes great contributions to the YOLOv8m model to achieve high-precision differentiation of four types of stress symptoms. In conclusion, our study provided an effective system to achieve low-cost, high-precision, fast, and infield diagnosis of tea stress symptoms in complex natural scenes based on smartphone and deep learning algorithms.

Ensuring the safety and efficacy of chemical compounds is crucial in small-molecule drug development. In the later stages of drug development, toxic compounds pose a significant challenge, losing valuable resources and time. Early and accurate prediction of compound toxicity using deep learning models offers a promising solution to mitigate these risks during drug discovery. In this study, we present the development of several deep-learning models aimed at evaluating different types of compound toxicity, including acute toxicity, carcinogenicity, hERG_cardiotoxicity (the human ether-a-go-go related gene caused cardiotoxicity), hepatotoxicity, and mutagenicity. To address the inherent variations in data size, label type, and distribution across different types of toxicity, we employed diverse training strategies. Our first approach involved utilizing a graph convolutional network (GCN) regression model to predict acute toxicity, which achieved notable performance with Pearson R 0.76, 0.74, and 0.65 for intraperitoneal, intravenous, and oral administration routes, respectively. Furthermore, we trained multiple GCN binary classification models, each tailored to a specific type of toxicity. These models exhibited high area under the curve (AUC) scores, with an impressive AUC of 0.69, 0.77, 0.88, and 0.79 for predicting carcinogenicity, hERG_cardiotoxicity, mutagenicity, and hepatotoxicity, respectively. Additionally, we have used the approved drug dataset to determine the appropriate threshold value for the prediction score in model usage. We integrated these models into a virtual screening pipeline to assess their effectiveness in identifying potential low-toxicity drug candidates. Our findings indicate that this deep learning approach has the potential to significantly reduce the cost and risk associated with drug development by expediting the selection of compounds with low toxicity profiles. Therefore, the models developed in this study hold promise as critical tools for early drug candidate screening and selection.

BACKGROUND: Chronic obstructive pulmonary disease (COPD) is underdiagnosed with the current gold standard measure pulmonary function test (PFT). A more sensitive and simple option for early detection and severity evaluation of COPD could benefit practitioners and patients.
METHODS: In this multicenter retrospective study, frontal chest X-ray (CXR) images and related clinical information of 1055 participants were collected and processed. Different deep learning algorithms and transfer learning models were trained to classify COPD based on clinical data and CXR images from 666 subjects, and validated in internal test set based on 284 participants. External test including 105 participants was also performed to verify the generalization ability of the learning algorithms in diagnosing COPD. Meanwhile, the model was further used to evaluate disease severity of COPD by predicting different grads.
RESULTS: The Ensemble model showed an AUC of 0.969 in distinguishing COPD by simultaneously extracting fusion features of clinical parameters and CXR images in internal test, better than models that used clinical parameters (AUC = 0.963) or images (AUC = 0.946) only. For the external test set, the AUC slightly declined to 0.934 in predicting COPD based on clinical parameters and CXR images. When applying the Ensemble model to determine disease severity of COPD, the AUC reached 0.894 for three-classification and 0.852 for five-classification respectively.
CONCLUSIONS: The present study used DL algorithms to screen COPD and predict disease severity based on CXR imaging and clinical parameters. The models showed good performance and the approach might be an effective case-finding tool with low radiation dose for COPD diagnosis and staging.

Color face images are often transmitted over public channels, where they are vulnerable to tampering attacks. To address this problem, the present paper introduces a novel scheme called Authentication and Color Face Self-Recovery (AuCFSR) for ensuring the authenticity of color face images and recovering the tampered areas in these images. AuCFSR uses a new two-dimensional hyperchaotic system called two-dimensional modular sine-cosine map (2D MSCM) to embed authentication and recovery data into the least significant bits of color image pixels. This produces high-quality output images with high security level. When tampered color face image is detected, AuCFSR executes two deep learning models: the CodeFormer model to enhance the visual quality of the recovered color face image and the DeOldify model to improve the colorization of this image. Experimental results demonstrate that AuCFSR outperforms recent similar schemes in tamper detection accuracy, security level, and visual quality of the recovered images.

Identifying critical safety management drivers with high driver-level risks is essential for traffic safety improvement. Previous studies commonly evaluated driver-level risks based upon aggregated statistical characteristics (e.g., driving exposure and driving behavior), which were obtained from long-period driving monitoring data. However, given the great advancements of the connected vehicle and in-vehicle data instrumentation technologies, there has been a notable increase in the collection of short-period driving data, which has emerged as a prominent data source for analysis. In this data environment, traditionally employed aggregated behavior characteristics are unstable due to the time-varying feature of driving behavior coupled with insufficient data sampling periods. Thus, traditional modeling methods based upon aggregated statistical characteristics are no longer feasible. Instead of utilizing such unreliable statistical information to represent driver-level risks, this study employed temporal variation characteristics of driving behavior to identify critical safety management drivers in the short-period driving data environment. Specifically, the relationships between driving behavior temporal variation characteristics and individual crash occurrence probability were developed. To eliminate the impacts of drivers\' driving behavior heterogeneity on model performance, \"traffic entropy\" index that could quantify the abnormal degrees of driving behavior was proposed. Deep learning models including convolutional neural network (CNN) and long short-term memory (LSTM) were employed to conduct the temporal variation feature mining. Empirical analyses were conducted using data obtained from online ride-hailing services. Experiment results showed that temporal variation characteristics based models outperformed traditional aggregated statistical characteristics based models. The area under the curve (AUC) index was improved by 4.1%. And the proposed traffic entropy index further enhanced the model performance by 5.3%. The best model achieved an AUC of 0.754, comparable to existing approaches utilizing long-period driving data. Finally, applications of the proposed method in driver management program development and its further investigations have been discussed.

In the realm of hyperspectral image classification, the pursuit of heightened accuracy and comprehensive feature extraction has led to the formulation of an advance architectural paradigm. This study proposed a model encapsulated within the framework of a unified model, which synergistically leverages the capabilities of three distinct branches: the swin transformer, convolutional neural network, and encoder-decoder. The main objective was to facilitate multiscale feature learning, a pivotal facet in hyperspectral image classification, with each branch specializing in unique facets of multiscale feature extraction. The swin transformer, recognized for its competence in distilling long-range dependencies, captures structural features across different scales; simultaneously, convolutional neural networks undertake localized feature extraction, engendering nuanced spatial information preservation. The encoder-decoder branch undertakes comprehensive analysis and reconstruction, fostering the assimilation of both multiscale spectral and spatial intricacies. To evaluate our approach, we conducted experiments on publicly available datasets and compared the results with state-of-the-art methods. Our proposed model obtains the best classification result compared to others. Specifically, overall accuracies of 96.87%, 98.48%, and 98.62% were obtained on the Xuzhou, Salinas, and LK datasets.

Toxic heavy gas sulfur dioxide (SO2) is a specific life and environmental hazard. Predicting the diffusion of SO2 has become a research focus in fields such as environmental and safety studies. However, traditional methods, such as kinetic models, cannot balance precision and time. Thus, they do not meet the needs of emergency decision-making. Deep learning (DL) models are emerging as a highly regarded solution, providing faster and more accurate predictions of gas concentrations. To this end, this study proposes an innovative hybrid DL model, the parallel-connected convolutional neural network-gated recurrent unit (PC CNN-GRU). This model utilizes two CNNs connected in parallel to process gas release and meteorological datasets, enabling the automatic extraction of high-dimensional data features and handling of long-term temporal dependencies through the GRU. The proposed model demonstrates good performance (RMSE, MAE, and R2 of 20.1658, 10.9158, and 0.9288, respectively) with real data from the Project Prairie Grass (PPG) case. Meanwhile, to address the issue of limited availability of raw data, in this study, time series generative adversarial network (TimeGAN) are introduced for SO2 diffusion studies for the first time, and their effectiveness is verified. To enhance the practicality of the research, the contribution of drivers to SO2 diffusion is quantified through the utilization of the permutation importance (PIMP) and Sobol\' method. Additionally, the maximum safe distance downwind under various conditions is visualized based on the SO2 toxicity endpoint concentration. The results of the analyses can provide a scientific basis for relevant decisions and measures.

\"Treatise on Febrile Diseases\" is an important classic book in the academic history of Chinese material medica. Based on the knowledge map of traditional Chinese medicine established by the study of \"Treatise on Febrile Diseases\", a question-answering system of traditional Chinese medicine was established to help people better understand and use traditional Chinese medicine. Intention classification is the basis of the question-answering system of traditional Chinese medicine, but as far as we know, there is no research on question intention classification based on \"Treatise on Febrile Diseases\". In this paper, the intent classification research is carried out based on the Chinese material medica-related content materials in \"Treatise on Febrile Diseases\" as data. Most of the existing models perform well on long text classification tasks, with high costs and a lot of memory requirements. However, the intent classification data of this paper has the characteristics of short text, a small amount of data, and unbalanced categories. In response to these problems, this paper proposes a knowledge distillation-based bidirectional Transformer encoder combined with a convolutional neural network model (TinyBERT-CNN), which is used for the task of question intent classification in \"Treatise on Febrile Diseases\". The model used TinyBERT as an embedding and encoding layer to obtain the global vector information of the text and then completed the intent classification by feeding the encoded feature information into the CNN. The experimental results indicated that the model outperformed other models in terms of accuracy, recall, and F1 values of 96.4%, 95.9%, and 96.2%, respectively. The experimental results prove that the model proposed in this paper can effectively classify the intent of the question sentences in \"Treatise on Febrile Diseases\", and provide technical support for the question-answering system of \"Treatise on Febrile Diseases\" later.