Deep Autoencoder Neural Network Research Articles

Optimizing electrical and thermal performance in AlGaN/GaN HEMT devices using dual‐metal gate technology

AbstractThe investigation of aluminum gallium nitride/gallium nitride high electron mobility transistor (AlGaN/GaN HEMT) devices with a dual‐metal gate (DMG) structure encompasses both electrical and thermal characteristics. As efforts to enhance heat dissipation progress, there is a concurrent exploration of novel semiconductor materials boasting high thermal conductivity, like boron arsenide and phosphide. Combining these materials into a model and measuring their interface achieves efficient energy transport. Minimizing the self‐heating impact in AlGaN/GaN HEMTs is essential for enhancing device efficiency. This research exposes the heterogenous combination of boron arsenide and phosphide cooling substrates with metals, GaN semiconductors and HEMT. In this research, the autoencoder deep neural network techniques in GaN HEMT for self‐heat reduction is driven by the ability to effectively analyze and model the thermal behavior of the device. Autoencoders learn complex relationships within temperature data and identify patterns associated with self‐heating. By leveraging these learned representations, the deep neural network optimizes control strategies to mitigate self‐heating effects in GaN HEMT devices, ultimately contributing to improved thermal management and enhanced overall performance. In this research, the use of genetic algorithms in GaN HEMT aims to optimize device parameters systematically, to minimize self‐heating effects and enhance overall thermal performance. The structure also enhances electron mobility within the channel. Results show DMG structures, exhibiting higher saturation output currents and transconductance despite self‐heating. The DMG exhibits a maximum gm value of 0.164 S/mm, which is 10% higher significantly enhancing GaN‐based HEMTs for improved reliability and efficiency in various applications.

Disruption detection for a cognitive digital supply chain twin using hybrid deep learning

Recent disruptive events, such as COVID-19 and Russia–Ukraine conflict, had a significant impact of global supply chains. Digital supply chain twins have been proposed in order to provide decision makers with an effective and efficient tool to mitigate disruption impact. This paper introduces a hybrid deep learning approach for disruption detection within a cognitive digital supply chain twin framework to enhance supply chain resilience. The proposed disruption detection module utilises a deep autoencoder neural network combined with a one-class support vector machine algorithm. In addition, long-short term memory neural network models are developed to identify the disrupted echelon and predict time-to-recovery from the disruption effect. The obtained information from the proposed approach will help decision-makers and supply chain practitioners make appropriate decisions aiming at minimizing negative impact of disruptive events based on real-time disruption detection data. The results demonstrate the trade-off between disruption detection model sensitivity, encountered delay in disruption detection, and false alarms. This approach has seldom been used in recent literature addressing this issue.

Bi-AAE: A binary adversarial autoencoder deep neural network model for anomaly detection in system-levels marine diesel engines

Marine diesel engines typically use multivariate time series for health condition monitoring, the anomaly detection of which is fundamental and critical for an entity's operation and management. However, detecting anomalies in multivariate time series remains a grand challenge due to the complex behavior of marine machinery. Therefore, this research proposes a binary adversarial autoencoder model to overcome this issue. The framework is built upon the principles of reconstruction models by fusing Generative Adversarial Networks and Adversarial Autoencoder approaches to analyze the degree of anomaly in multivariate time series. Additionally, it identifies anomalies through a threshold obtained by a statistical distribution-based and unsupervised setting method. To emphasize the model's performance, multivariate condition parameters from actual ship diesel engines are utilized for validation. Two membership functions are recognized as the model metrics to characterize its best performance, evidenced by scores of 0.937 for F1 scores and 0.910 for precision. The proposed strategy can be applied to different types of monitoring systems in the ship's engine room to realize system-level operation and maintenance.

Improving the second-tier classification of methylmalonic acidemia patients using a machinelearning ensemble method.

Methylmalonic acidemia (MMA) is a disorder of autosomal recessive inheritance, with an estimated prevalence of 1:50,000. First-tier clinical diagnostic tests often return many false positives [five false positive (FP): one true positive (TP)]. In this work, our goal was to refine a classification model that can minimize the number of false positives, currently an unmet need in the upstream diagnostics of MMA. We developed machinelearning multivariable screening models for MMA with utility as a secondary-tier tool for false positives reduction. We utilized mass spectrometry-based features consisting of 11 amino acids and 31 carnitines derived from dried blood samples of neonatal patients, followed by additional ratio feature construction. Feature selection strategies (selection by filter, recursive feature elimination, and learned vector quantization) were used to determine the input set for evaluating the performance of 14 classification models to identify a candidate model set for an ensemble model development. Our work identified computational models that explore metabolic analytes to reduce the number of false positives without compromising sensitivity. The best results [area under the receiver operating characteristic curve (AUROC) of 97%, sensitivity of 92%, and specificity of 95%] were obtained utilizing an ensemble of the algorithms random forest, C5.0, sparse linear discriminant analysis, and autoencoder deep neural network stacked with the algorithm stochastic gradient boosting as the supervisor. The model achieved a good performance trade-off for a screening application with 6% false-positive rate (FPR) at 95% sensitivity, 35% FPR at 99% sensitivity, and 39% FPR at 100% sensitivity. The classification results and approach of this research can be utilized by clinicians globally, to improve the overall discovery of MMA in pediatric patients. The improved method, when adjusted to 100% precision, can be used to further inform the diagnostic process journey of MMA and help reduce the burden for patients and their families.

Tracking element-mineral associations with unsupervised learning and dimensionality reduction in chemical and optical image stacks of thin sections

In-situ chemical analysis on thin sections is a cornerstone of geochemical research. Despite massive advances in digital image analysis, the interpretation of such data within the optical petrological context of the thin section has largely remained an analogue task in geochemistry. In this contribution, we registered optical and micro-chemical images from large thin section areas. Chemical datasets from scanning electron microscopy energy dispersive spectrometry (SEM-EDX) and Synchrotron X-ray fluorescence microscopy (S-XFM) were exported from proprietary software. We then evaluated two dimensionality reduction techniques against false-colour images and phase maps to test whether previously unnoticed features became discoverable. Principal component analysis (PCA) and deep sparse autoencoder (DSA) neural networks were used to summarise the multi-channel SEM-EDX and S-XFM into one single red-green-blue (RGB) representation each.Applied to an oceanic gabbro, a cratonic peridotite, and a pelagic limestone, the PCA-based dimensionality reduction was found to produce crisp RGB maps of phases with less noise than false-colour three-element RGB maps. They were registered with optical images to form a simple multi-channel input for pixel-based classification into classes (i.e., semantic segmentation). This was fast and worked well for typical igneous and metamorphic rocks (1–3 h routine). In the chemically quite homogenous limestone, the combined optical and S-XFM PCA input was successfully classified into many phases that were unclassifiable on the conventional SEM-EDX phase map. From the segmented optical and S-XFM PCA map, we exported pixel locations of classified accessory phases back into the original proprietary S-XFM software. This allowed quantification of trace elements whose X-ray peaks were invisible in the original S-XFM quantification.The DSA-based approach yielded phase maps with greater noise, but the technique was very strong at detecting subtle features. In the gabbro sample, the DSA image identified cryptic relict cores in clinopyroxene more clearly than in the previously used Cr concentration map. A DSA training set can be obtained on a small ROI to generate a consistent large area DSA image from which more fragmental cores were observed. For a ROI on the gabbro thin section, the DSA representation map and the blended optical image were segmented and inspected with a loupe tool. This allowed us to generate interactive histograms and bin-scatter concentration plots of selected phases and to search for co-variations of element concentrations. This approach can be expanded to include data from any in situ geochemical acquisition tools and may help discover previously unseen features in complex elemental imagery.

Health monitoring of an assembly-type structure using raw time-domain EMI signals and a novel hybrid deep learning algorithm

Deep learning models are widely used to extract features from data through supervised or unsupervised methods. However, when these two methods are used separately, their generalization ability is insufficient. In this study, a structural health monitoring technology based on a novel hybrid deep learning (HDL) model and time-domain electromechanical impedance (TEMI) has been proposed. The HDL is a highly efficient combination of deep auto-encoder (DAE) and two-dimensional convolutional neural network (2D CNN). The DAE is used for unsupervised reconstruction of the high-level features of 1D TEMI signals, and then, the features extracted by DAE are arranged and fed into a supervised 2D CNN that is designed to learn the optimal feature representations of the multi-label signals. This increases effectively the model’s learnability and generalizability. The robustness of the developed approach has been tested using TEMI signals that are collected from a scaled model of a rectangular pipe gallery subjected to transversal joint damage. Results indicate that the maximal identification errors of joint damage severity are 2.4% and 9.33% when the corresponding structural conditions are used and unused by the training process of HDL, respectively.

A review of recent advances in fault diagnosis based on deep neural networks

Bearings are essential components in mechanical systems, supporting the rotation of various machine parts in motors, wind turbines, vehicles, and industrial robots, making their health critical for system performance and reliability. Traditional diagnosis methods, such as vibration and acoustic analysis, along with temperature monitoring, often demand expertise and may struggle to detect early faults. However, the introduction of deep learning technology has created new opportunities for more effective bearing fault diagnosis. The application of deep learning-based bearing fault diagnosis in the industrial sector has gained significant attention and multiple types of deep learning networks have already been successfully implemented. This paper aims to provide a clear review of bearing fault diagnosis based on deep learning algorithms. This essay focuses on two of the most popular deep learning networks, Autoencoder and Convolutional Neural Networks. Their mechanism and applications are analyzed based on essays and research paper related to the field of bearing fault diagnosis. Finally, conclusions are presented to summarize the current development and point out faced challenges and future trends of these deep learning networks. It is also expected that this narrative not only serves as a cogent overview of the contemporary fault diagnosis technologies but also provides convenience and inspiration for further study in this field.

Utilizing Machine Learning to Examine the Spatiotemporal Changes in Africa’s Partial Atmospheric Layer Thickness

As a crucial aspect of the climate system, changes in Africa’s atmospheric layer thickness, i.e., the vertical distance spanning a specific layer of the Earth’s atmosphere, could impact its weather, air quality, and ecosystem. This study did not only examine the trends but also applied a deep autoencoder artificial neural network to detect years with significant anomalies in the thickness of Africa’s atmosphere over a given homogeneous region (derived with the rotated principal component analysis) and examine the fingerprint of global warming on the thickness changes. The broader implication of this study is to further categorize regions in Africa that have experienced significant changes in their climate system. The study reveals an upward trend in thickness between 1000 and 850 hPa across substantial parts of Africa since 1950. Notably, the spatial breadth of this rise peaks during the boreal summer. Correlation analysis, further supported by the deep autoencoder neural network, suggests the fingerprint of global warming signals on the increasing vertical extent of Africa’s atmosphere and is more pronounced (since the 2000s) in the south-central regions of Africa (specifically the Congo Basin). Additionally, the thickness over the Sahel and Sahara Desert sees no significant increase during the austral summer, resulting from the counteracting effect of the positive North Atlantic Oscillation, which prompts colder conditions over the northern parts of Africa. As the atmospheric layer thickness impacts the temperature and moisture distribution of the layer, our study contributes to its historical assessment for a sustainable ecosystem.

Robust optimization of geoenergy production using data-driven deep recurrent auto-encoder and fully-connected neural network proxy

Robust and efficient optimization of post-history well production schedule under history-matched geomodel known as closed-loop production management is crucial to achieve reasonable and reliable decision-making for subsurface geoenergy resources (e.g., oil/gas resources, geothermal, natural gas hydrate, etc.) recovery in petroleum industry. The procedure generally consists of a cycling implementation of geomodel history matching and production schedule optimization through repeatedly simulating high-fidelity reservoir models and therefore needs almost infeasible computation cost particularly for field-scale oil production scenario. Through hybridizing deep recurrent neural network and particle-swarm optimization algorithm, we present a novel data-driven computation framework for robust post-history production optimization without explicitly performing history matching. In the proposed framework, we develop a hybrid neural network proxy model, which combines recurrent autoencoder (RAE) and fully-connected neural network (FCNN), referred to as RAE-FCNN. The RAE structure can extract latent information from time-series history data, and the FCNN structure can map the combination of extracted latent variables by RAE and post-history well-control sequence to net present value (NPV). The trained RAE-FCNN proxy model can predict NPV values corresponding to the history measurements and user-specific post-history well controls. In the robust optimization framework, the averaged NPV over an ensemble of noisy history measurements are regarded as the objective function while the post-history well-controls are chosen as the decision variables. The proposed surrogate-based optimization framework has been demonstrated on two waterflooding reservoir production cases, e.g., a 2D synthetic Gaussian model and a 3D benchmaRk channelized EGG model.The decision variables considered in optimization are the injection rates with linear bound constraints at the injectors. In comparison with the optimization results by high-fidelity reservoir models, solely 1000 high-fidelity reservoir models for both cases are sufficient to train accurate proxy model and achieve satisfactory post-history production optimization results. Our proposed deep-learning-based optimization method can effectively make full use of history data to predict post-history oil well production without the need of computationally expensive history matching step, which enables us to achieve quick decision-making and risk evaluation for close-loop waterflooding reservoir production management and optimization.

Intrusion detection in big data environment using hybrid deep learning algorithm (VAE-CNN)

In the internet era, billions of devices are connected to the network generates large volume of data and the generation rate increases exponentially every day. As the data increases, the chances for cyber attackers to exploit the data increases which results into numerous security threats to organizations and network. Fast and accurate detection of attacks in big data environment is difficult due to its volume and variety and velocity. Over a decade, numerous attack detection systems are developed using machine learning. However, most of the traditional detection systems cannot recognize the attack types specifically which reduces the detection performances and network performances. Thus, the intrusion detection model presented in this research which incorporates deep variational auto-encoder and convolutional neural network to detect intrusions. Experimentations using benchmark dataset validated the proposed model better performances over existing machine learning techniques like logistic regression, random forest, extreme gradient boosting, k-nearest neighbor, and self-scalable heuristic artificial neural network algorithms using accuracy, recall, precision, and F1-score. The proposed model outperforms with a maximum precision of 97.48%, Recall of 99.52%, F1-score of 98.49% and accuracy of 98.65% over conventional intrusion detection algorithms.

Recognition of mineralization-related anomaly patterns through an autoencoder neural network for mineral exploration targeting

In mineral potential mapping, supervised machine learning algorithms have shown great promise in delineating and prioritizing potential areas. However, since mineralization being a relatively rare geological event, most supervised machine learning-based models face substantial challenges in properly identifying prospective areas. Data sets with strongly imbalanced distributions of the target variable (deposits) and insufficient training data sets impose obstacles to these kinds of models which can significantly impact adversely on the performance of the models. Moreover, in some cases, negative training data sets as the non-deposit locations aren't really true negative data, which cause higher uncertainty in a mineral potential map. In this study, for handling these challenges the deep autoencoder neural network is adopted. The autoencoder can be trained to reconstruct geospatial data set in totally unsupervised manner and identify prospective areas based on the reconstruction error, where higher error corresponds with areas of higher mineral potential. In order to confirm the efficiency of the autoencoder algorithm in mineral potential modeling, the model was compared with a popular data-driven approach that assigned a weight to the evidence layer by using a concentration-area (C-A) fractal model and a prediction-area (P-A) plot, and combined them using a multi-class index overlay method. Receiver operating characteristic (ROC) curve, success-rate curve, and P-A plot were adopted to evaluate the predictive ability of Fe prospectivity models pertaining to the Esfordi district of Iran. Also, we use an area under the ROC curve (AUC) and partial AUC (pAUC) to quantitatively evaluate the overall and sensitivity performance of models, respectively.

A deep learning approach for anomaly identification in PZT sensors using point contact method

The implementation of piezoelectric sensors is degraded due to surface defects, delamination, and extreme weathering conditions, to mention a few. Hence, the sensor needs to be diagnosed before the efficacious implementation in the structural health monitoring (SHM) framework. To rescue the problem, a novel experimental method based on Coulomb coupling is utilised to visualise the evolution of elastic waves and interaction with the surface anomaly in the lead zirconate titanate (PZT) substrate. Recently, machine learning (ML) has been expeditiously becoming an essential technology for scientific computing, with several possibilities to advance the field of SHM. This study employs a deep learning-based autoencoder neural network in conjunction with image registration and peak signal-to-noise ratio (PSNR) to diagnose the surface anomaly in the PZT substrate. The autoencoder extracts the significant damage-sensitive features from the complex waveform big data. Further, it provides a nonlinear input–output model that is well suited for the non-linear interaction of the wave with the surface anomaly and boundary of the substrate. The measured time-series waveform data is provided as input into the autoencoder network. The mean absolute error (MAE) between the input and output of the deep learning model is evaluated to detect the anomaly. The MAEs are sensitive to the anomaly that lies in the PZT substrate. Further, the challenge arising from offset and distortion is addressed with ad hoc image registration technique. Finally, the localisation and quantification of the anomaly are performed by computing PSNR values. This work proposes an advanced, efficient damage detection algorithm in the scenario of big data that is ubiquitous in SHM.

Intelligent IoT Enabled Pipeline Leakage Detection and Monitoring System Using Machine Learning Algorithm

Pipeline networks are the safest transportation of oil and gas-related products that may be susceptible to failures. These kinds of failures are due to the reason of manufacturing defects, degradation of the material, environmental defects, and interference of the third party. These issues are addressed through the Internet of Things (IoT) based solutions and give promising outcomes in predicting the failures and monitoring. The IoT can combine the sensor technologies to monitor the gas flow, pressure, temperature, condition of the compressor, concentration, and other related features inside the pipeline. The leaks in the pipe are measured as a pin-sized hole that is detected by the IoT sensor with advanced technologies such as deep learning. This project proposed deep learning with a meta- heuristic approach-based model to detect leaks in pipelines through IoT sensors. The Deep Learning model called Deep Auto Encoder Neural network (DAENN) is the unsupervised model that can accurately classify the leaking and non-leaking pipeline conditions. The detection accuracy is further enhanced with the Bat Optimization algorithm (BOA) which obtained improved accuracy while leaks occur in the pipeline inside the sensor monitoring area. This observation deploys the monitoring sensors to cover the mentioned monitoring area. The proposed model can increase the system’s leak detection reliability and reduces the false alarm rate.

Towards reliable hepatocytic anatomy segmentation in laparoscopic cholecystectomy using U-Net with Auto-Encoder.

Most bile duct (BDI) injuries during laparoscopic cholecystectomy (LC) occur due to visual misperception leading to the misinterpretation of anatomy. Deep learning (DL) models for surgical video analysis could, therefore, support visual tasks such as identifying critical view of safety (CVS). This study aims to develop a prediction model of CVS during LC. This aim is accomplished using a deep neural network integrated with a segmentation model that is capable of highlighting hepatocytic anatomy. Still images from LC videos were annotated with four hepatocystic landmarks of anatomy segmentation. A deep autoencoder neural network with U-Net to investigate accurate medical image segmentation was trained and tested using fivefold cross-validation. Accuracy, Loss, Intersection over Union (IoU), Precision, Recall, and Hausdorff Distance were computed to evaluate the model performance versus the annotated ground truth. A total of 1550 images from 200 LC videos were annotated. Mean IoU for segmentation was 74.65%. The proposed approach performed well for automatic hepatocytic landmarks identification with 92% accuracy and 93.9% precision and can segment challenging cases. DL, can potentially provide an intraoperative model for surgical video analysis and can be trained to guide surgeons toward reliable hepatocytic anatomy segmentation and produce selective video documentation of this safety step of LC.

Automated Tongue Diagnosis: A Deep Autoencoder Neural Network and Clustering-Based Image Segmentation Approach

Automated tongue diagnosis has gained significant attention in recent years as a non-invasive and cost-effective method for disease detection and monitoring. In this project, propose a novel approach for automated tongue diagnosis using a combination of Deep Autoencoder Neural Network (DAENN) and clustering-based image segmentation techniques. The objective is to develop a reliable and efficient system that can accurately analyze tongue images and provide diagnostic information. The proposed methodology involves several key steps. First, high-resolution images of the tongue are acquired using a digital camera. Preprocessing techniques are then applied to enhance image quality and remove noise. Next, a clustering-based image segmentation algorithm is employed to extract the tongue region from the background. This step ensures that subsequent analysis is focused on the relevant area of interest. After tongue region extraction, relevant features are extracted from the segmented image. These features capture important visual characteristics such as color distribution, texture, and shape. The DAENN is then trained using these features to learn an efficient representation of the input data. The network's encoder-decoder architecture enables it to compress the features into a lower-dimensional latent space and reconstruct the original features. By leveraging advanced machine learning techniques, the system can accurately analyze tongue images and assist healthcare professionals in making informed decisions. The evaluation of the proposed system on a diverse dataset demonstrates promising results, showcasing its potential as an effective tool for automated tongue diagnosis. Keywords—deep autoencoder, Tongue segmentation, Pattern recognition, colour intensity

Predicting potential microbe-disease associations based on multi-source features and deep learning.

Studies have confirmed that the occurrence of many complex diseases in the human body is closely related to the microbial community, and microbes can affect tumorigenesis and metastasis by regulating the tumor microenvironment. However, there are still large gaps in the clinical observation of the microbiota in disease. Although biological experiments are accurate in identifying disease-associated microbes, they are also time-consuming and expensive. The computational models for effective identification of diseases related microbes can shorten this process, and reduce capital and time costs. Based on this, in the paper, a model named DSAE_RF is presented to predict latent microbe-disease associations by combining multi-source features and deep learning. DSAE_RF calculates four similarities between microbes and diseases, which are then used as feature vectors for the disease-microbe pairs. Later, reliable negative samples are screened by k-means clustering, and a deep sparse autoencoder neural network is further used to extract effective features of the disease-microbe pairs. In this foundation, a random forest classifier is presented to predict the associations between microbes and diseases. To assess the performance of the model in this paper, 10-fold cross-validation is implemented on the same dataset. As a result, the AUC and AUPR of the model are 0.9448 and 0.9431, respectively. Furthermore, we also conduct a variety of experiments, including comparison of negative sample selection methods, comparison with different models and classifiers, Kolmogorov-Smirnov test and t-test, ablation experiments, robustness analysis, and case studies on Covid-19 and colorectal cancer. The results fully demonstrate the reliability and availability of our model.

Pilot Study on the Use of Untargeted Metabolomic Fingerprinting of Liquid-Cytology Fluids as a Diagnostic Tool of Malignancy for Thyroid Nodules.

Although it is the gold standard for assessing the malignancy of thyroid nodules (TNs) preoperatively, the cytological analysis of fine-needle aspiration cytology (FNAC) samples results in 20-30% of cases in indeterminate lesions (ITNs). As two-thirds of these lesions will appear benign after diagnostic surgery, improved preoperative diagnostic methods need to be developed. In this pilot study, we evaluate if the metabolomic profiles of liquid-based (CytoRich®) FNAC samples of benign and malignant nodules can allow the molecular diagnosis of TNs. We performed untargeted metabolomic analyses with CytoRich® FNAC in a monocentric retrospective study. The cohort was composed of cytologically benign TNs, histologically benign or papillary thyroid carcinomas (PTCs) cytologically ITNs, and suspicious/malignant TNs histologically confirmed as PTCs. The diagnostic performance of the identified metabolomic signature was assessed using several supervised classification methods. Seventy-eight patients were enrolled in the study. We identified 7690 peaks, of which 2697 ions were included for further analysis. We selected a metabolomic signature composed of the top 15 metabolites. Among all the supervised classification methods, the supervised autoencoder deep neural network exhibited the best performance, with an accuracy of 0.957 (0.842-1), an AUC of 0.945 (0.833-1), and an F1 score of 0.947 (0.842-1). Here, we report a promising new ancillary molecular technique to differentiate PTCs from benign TNs (including among ITNs) based on the metabolomic signature of FNAC sample fluids. Further studies with larger cohorts are now needed to identify a larger number of biomarkers and obtain more robust signatures.

Railway Freight Demand Forecasting Based on Multiple Factors: Grey Relational Analysis and Deep Autoencoder Neural Networks

The construction of high-speed rail lines in China has drastically improved the freight capacity of conventional railways. However, due to recent national energy policy adjustments, rail freight volumes, consisting mostly of coal, ore, and other minerals, have declined. As a result, the corresponding changes in the supply and demand of goods and transportation have led to a gradual transformation of the railway freight market from a seller’s market to a buyer’s market. It is important to carry out a systematic analysis and a precise forecast of the demand for rail freight transport. However, traditional time series forecasting models often lack precision during drastic fluctuations in demand, while deep learning-based forecasting models may lack interpretability. This study combines grey relational analysis (GRA) and deep neural networks (DNN) to offer a more interpretable approach to predicting rail freight demand. GRA is used to obtain explanatory variables associated with railway freight demand, which improves the intelligibility of the DNN prediction. However, the high-dimension predictor variable can make training on DNN challenging. Inspired by deep autoencoders (DAE), we add a layer of an encoder to the GRA-DNN model to compress and aggregate the high-dimension input. Case studies conducted on Chinese railway freight from 2000 to 2018 show that the proven GRA-DAE-NN model is precise and easy to interpret. Comparative experiments with conventional prediction models ARIMA, SVR, FC-LSTM, DNN, FNN, and GRNN further validate the performance of the GRA-DAE-NN model. The prediction accuracy of the GRA-DAE-NN model is 97.79%, higher than that of other models. Among the main explanatory variables, coal, oil, grain production, railway locomotives, and vehicles have a significant impact on the railway freight demand trend. The ablation experiment verified that GRA has a significant effect on the selection of explanatory variables and on improving the accuracy of predictions. The method proposed in this study not only accurately predicts railway freight demand but also helps railway transportation companies to better understand the key factors influencing demand changes.

Asphalt Concrete Mix Design Optimization Using Autoencoder Deep Neural Networks

Asphalt concrete (AC) balanced mix design (BMD) is based on the selection of aggregate gradation, component volumetrics, and binder content to control pavement cracking and rutting potential. The Illinois Flexibility Index Test (I-FIT) and the Hamburg Wheel Tracking Test (HWTT) results, used to predict cracking and rutting potential, respectively, are used in the BMD approach. However, BMD generally relies on a trial-and-error process to identify the aggregate gradation and binder content needed to meet volumetrics and optimize I-FIT and HWTT results. Minimizing or eliminating the trial-and-error process would increase productivity and accuracy. Therefore, this study proposes an autoencoder deep neural network (ADNN) to develop optimized AC mix design alternatives that can meet a prescribed flexibility index (FI) and rut depth (RD). Autoencoders are a type of neural network designed for representation learning composed of an encoder and a decoder. The encoder detects a structured pattern in the original input data to create a compressed representation of the AC mix design. The decoder reconstructs the compressed representation. The proposed autoencoder is composed of an encoder of five hidden layers, a latent space of one node, and a five-hidden-layer decoder. Models were created from a database of 5,357 data sets that include mix properties, I-FIT FI, and HWTT RD (after data preprocessing was conducted). An autoencoder was then trained to predict the total binder content, and aggregate gradation based on a target mix type, FI, and RD.

IoT botnet attack detection using deep autoencoder and artificial neural networks

IoT botnet attack detection using deep autoencoder and artificial neural networks