Network Model Research Articles

Small-scale cross-layer fusion network for classification of diabetic retinopathy

Deep learning-based automatic classification of diabetic retinopathy (DR) helps to enhance the accuracy and efficiency of auxiliary diagnosis. This paper presents an improved residual network model for classifying DR into five different severity levels. First, the convolution in the first layer of the residual network was replaced with three smaller convolutions to reduce the computational load of the network. Second, to address the issue of inaccurate classification due to minimal differences between different severity levels, a mixed attention mechanism was introduced to make the model focus more on the crucial features of the lesions. Finally, to better extract the morphological features of the lesions in DR images, cross-layer fusion convolutions were used instead of the conventional residual structure. To validate the effectiveness of the improved model, it was applied to the Kaggle Blindness Detection competition dataset APTOS2019. The experimental results demonstrated that the proposed model achieved a classification accuracy of 97.75% and a Kappa value of 0.971 7 for the five DR severity levels. Compared to some existing models, this approach shows significant advantages in classification accuracy and performance.

A multi-step regularity assessment and joint prediction system for ordering time series based on entropy and deep learning

Customer maintenance is of vital importance to the enterprise management. Valuable assessment and efficient prediction for customer ordering behavior can offer better decision-making and reduce business costs significantly. According to existing studies about customer behavior regularity segment and demand prediction most focus on e-commerce and other fields with large amount of data, making them not suitable for small enterprises and data features like sparsity and outliers are not mined when doing regularity quantification. Additionally, more and more complex network structures for demand prediction are proposed, which builds on the assumption that all the samples have predictive value, ignoring the fine-grained analysis of different time series regularity with high cost. To deal with the above issues, a multi-step regularity assessment and joint prediction system for ordering time series is proposed. For extracting features, comprehensive assessment of customer regularity based on entropy weight method with the result of predictability quantification using K-Means clustering algorithm, real entropy, LZW algorithm and anomaly detection adopting Isolation Forest algorithm not only gives an objective result to ‘how high the regularity of customers is’, filling the gap in the field of regularity quantification, but also provides a theoretical basis for demand prediction models selection. Prediction models: Random Forest regression, XGBoost, CNN and LSTM network are experimented with sMAPE and MSLE for performance evaluation to verify the effectiveness of the proposed regularity quantitation method. Moreover, a merged CNN-BiLSTM neural network model is established for predicting those customers with low regularity and difficult to predict by traditional machine leaning algorithms, which performs better on the data set compared to others. Random Forest is still used for prediction of customers with high regularity due to its high training efficiency. Finally, the results of prediction, regularity quantification, and classification are output from the intelligent system, which is capable of providing scientific basis for corporate strategy decision and has highly extendibility in other enterprises and fields for follow-up research.

Noninvasive Deep Learning Analysis for Smith–Magenis Syndrome Classification

Smith–Magenis syndrome (SMS) is a rare, underdiagnosed condition due to limited public awareness of genetic testing and a lengthy diagnostic process. Voice analysis can be a noninvasive tool for monitoring and detecting SMS. In this paper, the cepstral peak prominence and mel-frequency cepstral coefficients are used as disease monitoring and detection metrics. In addition, an efficient neural network, incorporating synthetic data processes, was used to detect SMS in a cohort of individuals with the disease. Three study cases were conducted with a set of 19 SMS patients and 292 controls. The three study cases employed various oversampling and undersampling techniques, including SMOTE, random oversampling, NearMiss, random undersampling, and 16 additional methods, resulting in balanced accuracies ranging from 69% to 92%. This is the first study using a neural network model to focus on a rare genetic syndrome using phonation analysis data. By using synthetic data (oversampling and undersampling) and a CNN, it was possible to detect SMS with high levels of accuracy. Voice analysis and deep learning techniques have proven to be a useful and noninvasive method. This is a finding that may help in the complex identification of this syndrome as well as other rare diseases.

Study on Compression Bearing Capacity of Tapered Concrete-Filled Double-Skin Steel Tubular Members Based on Heuristic-Algorithm-Optimized Backpropagation Neural Network Model

A tapered concrete-filled double-skin steel tubular (TCFDST) structure has been used as the main framework in transmission towers, offshore facility platforms, and turbine towers owing to its excellent mechanical properties. In order to solve the difficulties of calculating the axial compressive capacity of TCFDST members due to the variations in cross-section, this paper applied heuristic optimization algorithms such as Genetic Algorithms (GAs), Particle Swarm Optimization (PSO), Simulated Annealing (SA), and Ant Colony Optimization (ACO) to enhance a Backpropagation Neural Network (BPNN) model. A predictive model incorporating both global and local optimization strategies for the axial compressive capacity of a TCFDST structure is proposed. A comprehensive axial database for TCFDST members, comprising 1327 sets of experimental and finite element analysis results, was established, with ten types of component dimensions and material parameters selected as input variables and compressive bearing capacity as the output variable. This study developed and assessed four BPNN models, each optimized by a different heuristic algorithm, against various machine learning algorithms and standards. The heuristic-algorithm-optimized BPNN models demonstrated superior accuracy in predicting the axial compressive capacity of TCFDST members. Through parametric analysis, this study identified the relationship between the model’s bearing capacity predictions and each input parameter, confirming the model’s broad applicability. The optimized BPNN model, refined with heuristic algorithms, provides a significant reference for addressing the computational challenges associated with the load-bearing capacity of TCFDST structures and facilitating their application.

Deep learning assisted cancer disease prediction from gene expression data using WT-GAN

Several diverse fields including the healthcare system and drug development sectors have benefited immensely through the adoption of deep learning (DL), which is a subset of artificial intelligence (AI) and machine learning (ML). Cancer makes up a significant percentage of the illnesses that cause early human mortality across the globe, and this situation is likely to rise in the coming years, especially when non-communicable illnesses are not considered. As a result, cancer patients would greatly benefit from precise and timely diagnosis and prediction. Deep learning (DL) has become a common technique in healthcare due to the abundance of computational power. Gene expression datasets are frequently used in major DL-based applications for illness detection, notably in cancer therapy. The quantity of medical data, on the other hand, is often insufficient to fulfill deep learning requirements. Microarray gene expression datasets are used for training procedures despite their extreme dimensionality, limited volume of data samples, and sparsely available information. Data augmentation is commonly used to expand the training sample size for gene data. The Wasserstein Tabular Generative Adversarial Network (WT-GAN) model is used for the data augmentation process for generating synthetic data in this proposed work. The correlation-based feature selection technique selects the most relevant characteristics based on threshold values. Deep FNN and ML algorithms train and classify the gene expression samples. The augmented data give better classification results (> 97%) when using WT-GAN for cancer diagnosis.

XElemNet: towards explainable AI for deep neural networks in materials science

Recent progress in deep learning has significantly impacted materials science, leading to accelerated material discovery and innovation. ElemNet, a deep neural network model that predicts formation energy from elemental compositions, exemplifies the application of deep learning techniques in this field. However, the “black-box” nature of deep learning models often raises concerns about their interpretability and reliability. In this study, we propose XElemNet to explore the interpretability of ElemNet by applying a series of explainable artificial intelligence (XAI) techniques, focusing on post-hoc analysis and model transparency. The experiments with artificial binary datasets reveal ElemNet’s effectiveness in predicting convex hulls of element-pair systems across periodic table groups, indicating its capability to effectively discern elemental interactions in most cases. Additionally, feature importance analysis within ElemNet highlights alignment with chemical properties of elements such as reactivity and electronegativity. XElemNet provides insights into the strengths and limitations of ElemNet and offers a potential pathway for explaining other deep learning models in materials science.

Pore-scale binary diffusion behavior of Hydrogen-Cushion gas in saline aquifers for underground hydrogen Storage: Optimization of cushion gas type

Cushion gas is required to prevent hydrogen (H2) trapping and interactions between H2 and brine for underground hydrogen storage (UHS) in saline aquifers. However, selecting an appropriate type of cushion gas is not a straightforward task, as it affects the H2 diffusion characteristic in subsurface porous media, thereby impacting purity, recoverability, etc. In this paper, we developed a modified pore network model (PNM) for binary diffusion that incorporates multi-scale diffusion mechanisms to forecast the H2 effective diffusion coefficient (DH2e) and evaluate the impacts of the cushion gas on DH2e. Additionally, the model accounts for the water-lock effects of residual brine, focusing exclusively on the connected pore-throats occupied by the gas phase, which are identified using the invasion percolation algorithm with trapping. Our key findings are as follows: 1) CO2 and N2 demonstrate superior containment effects on H2 compared to CH4 under the specific conditions of the target aquifer, reducing the H2 contamination caused by binary diffusion; 2) The size and topology of porous media exert minimal impacts on the selection of cushion gas, yet they markedly affect the DH2e; 3) The pressure and temperature conditions of the saline aquifer are critical factors in cushion gas selection, with temperature being the more influential parameter. This study provides valuable insights for the implementation of industry-scale UHS in saline aquifers.

VSegNet - A Variant SegNet for Improving Segmentation Accuracy in Medical Images with Class Imbalance and Limited Data

Deep learning methods for many medical image segmentation task encounter challenges like smaller datasets and class imbalance. This study proposes a variant SegNet (vSegNet) designed to deliver significantly accurate and reliable segmentation results on such datasets. The novelty lies in designing encoder and decoder blocks with an appropriate number of convolution layers and using the Dice score and Hausdorff distance (HD) as compound loss function in learning. This study used public datasets consisting of chest X-rays, axial CT slices, foot ulcer images, and subset of SPIDER dataset to benchmark segmentation task of the proposed neural network model with other popular networks like U-Net, SegNet, DeepLabv3+, VGG16, MobileNetV2, and fully convolutional network (FCN). For the segmentation of lungs in chest X-rays, vertebral body in CT, augmented data for the previous case, foot ulcer dataset, and segmentation of vertebrae, intervertebral disks and spinal canal in SPIDER dataset (MRI dataset) respectively, the proposed vSegNet performed with a Dice score of 0.96 ± 0.01, 0.90 ± 0.20, 0.95 ± 0.02, 0.86 ± 0.07, and 0.95 ± 0.01 and the HD of 14.33 ± 7.74, 8.45 ± 7.08, 7.99 ± 6.05, 29.32 ± 25.64, and 8.45 ± 2.81 with respect to the ground truth on the test dataset. These results highlight the effectiveness of the proposed model in delivering both higher segmentation accuracy and improved boundary delineation. The proposed network, vSegNet has demonstrated as an effective model for semantic segmentation on class-imbalanced smaller datasets, surpassing all other networks considered in this study in terms of mIoU, BF score, Dice score, HD, accuracy, precision, recall and F1 score on a variety of anatomical regions and medical imaging modalities.

Fractional Flow Analysis of Foam Displacement in Tight Porous Media with Quasi-Static Pore Network Modeling and Core-Flooding Experiments

Fractional flow analysis is an efficient tool to evaluate the gas-trapping performance of foam in porous media. The pore-scale simulation study and the core-scale experimental work have been bridged via the fractional flow analysis to distinguish the characteristics of foam displacement inside the tight porous media with varying absolute permeability, injection rate, and foam quality. In this work, the combined investigation suggests that conventional foam-enhancing strategies, pursuing higher foam quality and stronger foam regime, are inefficient and restricted in tight reservoirs that the critical Sw corresponding to the limiting capillary pressure has increased around 37~43%, which indicates severely weakened gas-trapping capacity as permeability reduces one order of magnitude. The moderate mobility adjustment and corresponding optimized fluid injectivity exerting from the “weak foam” flow presents a staged decline feature of decreasing water fractional flow, which implies the existence of the delayed gas-trapping phenomenon when water saturation reduces to 0.5~0.6. The finding has supported the engineering ideal of promoting low-tension gas (LTG) drive processes as a potential solution to assist field gas injection applications suffering from gas channeling. Also, the validation with core-flooding experimental results has revealed several defects of the current pore network model of foam displacement in tight porous media, including exaggerated gas trapping and overestimated confining water saturation. This study has innovatively demonstrated the feasibility and potential of optimizing the foam performance of gas trapping and mobility control in tight reservoirs, which provides a clue that may eventually boost the efficiency of the gas injection process in enhanced oil recovery or CO2 sequestration projects.

Fast acquisition method of battery electrochemical impedance spectra based on impedance fragments

Electrochemical impedance spectroscopy (EIS) is often used for battery characterization and monitoring. However, there are still difficulties in acquiring battery impedance spectra in large quantities and wide frequency bands in practical applications. The lower the frequency range of acquisition, the longer the measurement time spent, and the higher the frequency range of measurement, the more expensive the measurement equipment, and these problems seriously limit the wide application of impedance information. To address this problem, this study proposes a method for fast acquisition of electrochemical impedance spectra of lithium-ion batteries based on impedance fragments. Firstly, impedance segments at specific frequencies in the EIS are selected as features based on the distribution of relaxation times (DRT). Then, a Long Short-Term Memory (LSTM) neural network model is used to reconstruct the complete impedance spectrum, and a population optimization algorithm is used to optimize the hyperparameters in the model. The validation results indicate that the reconstructed RMSE error is only 14.26 mΩ, corresponding to a MAPE of 1.65 %, while the measurement time is reduced to 27.01 % of the original measurement time. Finally, the performance of this method in identifying fractional-order model parameters is further discussed.

Efficient Fault Detection and Analysis of Power System Distribution Networks by Integrating BP Data Mining

As the basic guarantee for people’s production and life, the safe operation of the power system has an important impact on the development and operation of society. To ensure the safe and stable operation of the power grid, predicting potential faults and taking reasonable preventive measures can effectively avoid the occurrence of power accidents. However, due to the difficulty in ensuring the prediction accuracy of traditional methods, there are issues of protection misoperation and rejection. Therefore, in order to achieve accurate prediction of power grid faults and avoid protection misoperation and rejection issues, a distribution network fault classification prediction model using a combination of three-layer data mining model (TLDM) and adaptive moment estimation (Adam) algorithm/random gradient descent algorithm improved backpropagation neural network (BPNN) is proposed. The implementation results showed that the classification accuracy of artificial fish school [Formula: see text] priori, [Formula: see text]-means clustering convolutional neural network model and TLDM for single-phase grounding faults was 93.2%, 91.5% and 96.6%, respectively. The classification accuracy for two-phase faults was 92.8%, 92.4% and 95.7%, respectively. The classification accuracy for two-phase grounding faults was 93.7%, 91.2% and 96.9%, respectively. The classification accuracy for three-phase faults was 93.3%, 92.1% and 97.1%, respectively. The TLDM had the highest classification accuracy. The average accuracy, average accuracy and average recall of the BPNN improved by the combination of the ADAM algorithm and random gradient descent algorithm were 94.1%, 90.9% and 88%, respectively, which were higher than the BPNN improved by the combination of ADAM algorithm and random gradient descent algorithm. The above results indicate that the proposed distribution network fault classification and prediction model has good performance and can achieve accurate prediction of distribution network faults.

Fire Recognition Method Based on PSO-BP Neural Network and ResNet50

The rapid development of modern society and continuous urbanization have resulted in a proliferation of functional buildings, which offer significant convenience to individuals, but pose significant fire hazards as well. How to detect the fire at the early stage is always the focus of research. This paper proposes a multi-information source fusion fire recognition method based on particle swarm optimization (PSO)-backpropagation (BP) neural networks and ResNet50. The PSO algorithm is applied to optimize the initial parameters of a BP neural network model, while data from three sensors — temperature, humidity and smoke — are integrated, through iterative training of the system, accurate recognition of sensor data can be achieved. Additionally, a method is proposed for the recognition of infrared fire images using ResNet50 and transfer learning. By improving the ResNet50 network model and migrating the ResNet50 pre-trained network weight, infrared fire image recognition accuracy is further enhanced. Then the sensor information recognition results and image information recognition results are input into the fuzzy system for fusion reasoning again, and the final decision is output according to the set fuzzy rules. Experimental findings demonstrate that the multi-information source fusion approach utilizing the PSO-BP neural network and ResNet50 significantly enhances the accuracy and response time of fire recognition, and achieves a remarkable recognition effect.

An investigation of the Bernstein’s strengths Scale: factorial validity and network analysis of attention-deficit/hyperactivity symptoms, mental health, and the strengths of the healthy adult self

BackgroundBernstein’s heuristic model of personal strengths describes 16 positive attributes of the healthy adult self, grouped into four higher-order factors: self-directedness (identity, self-reflection, self-confidence, self-assertion, imagination/creativity), self-regulation (emotional balance, resilience, self-control, self-care, reality testing), connection (empathy, compassion, humour, responsibility), and transcendence (gratitude and wisdom). The study aimed to explore the factorial validity of Bernstein’s Strengths Scale developed for the assessment of these constructs. Furthermore, we explored the association of strengths with attention-deficit hyperactivity (ADHD) symptoms and mental health in a nonclinical sample using a network approach.MethodsFive hundred and twenty-eight adults filled out an online questionnaire, including Bernstein’s Strengths Scale (BSS), the Mental Health Continuum – Short Form (MHC-SF), and the Adult ADHD Self-Report Scale (ASRS).ResultsThough our data showed an acceptable fit to the 16-factor model, the Self-control subscale and item 28 of the Compassion subscale were excluded because of reliability issues. The final model with 15 factors and 44 items showed excellent fit, χ2(797) = 1130.149, p < .001, RMSEA = 0.028, 90% CI [0.024, 0.032], SRMS = 0.049, CFI = 0.993, TLI = 0.991. Principal component analysis revealed three instead of four second-order factors. In the network model, Self-confidence and Wisdom emerged as the most influential nodes. ADHD symptoms were negatively connected to Responsibility, Self-care, Identity, Emotional balance, Self-confidence, and mental health, while they had positive edges with Humor, Self-reflection, and Imagination/creativity. Mental health was positively related to several strengths, with the strongest association being between mental health and Gratitude.ConclusionThe BSS has acceptable factorial validity and is a reliable tool for assessing the personal strengths of the healthy adult self. ADHD symptoms in adults may be associated with deficits in personal strengths, but protective factors could also be identified. Addressing personal strengths in psychosocial interventions for ADHD in adults may improve patients’ emotional well-being and everyday functioning.

Modality deep-learning frameworks for fake news detection on social networks: a systematic literature review

Fake news on social networks is a challenging problem due to the rapid dissemination and volume of information, as well as the ease of creating and sharing content anonymously. Fake news stories are problematic not only for the credibility of online journalism, but also due to their detrimental real-world consequences. The primary research objective of this study is: What are the recent state-of-the-art modalities based on deep learning to detect fake news in social networks. This paper presents a systematic literature review of deep learning-based fake news detection models in social networks. The methodology followed a rigorous approach, including predefined criteria for study selection of deep learning modalities. This study focuses on the types of deep learning modalities; unimodal (refers to the use of a single model for analysis or modeling purposes) and multimodal models (refers to the integration of multiple models). The results of this review reveal the strengths and weaknesses of modalities approaches, as well as the limitations of low-resource languages datasets. Furthermore, it provides insights into future directions for deep learning models and different fact checking techniques. At the end of this study, we discuss the problem of fake news detection in the era of large language models in terms of advantages, drawbacks, and challenges.

The interchange stations siting model of urban rail transit network

The "node anchor weaving method" is commonly used in urban rail transit network planning, which emphasizes identifying the important nodes of the network first, and then using the key nodes as the "anchor nodes" to construct the network. Interchange stations are the most crucial network nodes in terms of both service function and network function. As a result, when planning the network, it is critical to identify the interchanges ahead of time. A key node identification model is initially established in this article, and then a network interchange stations location model is built on top of it. Based on the urban road network, the key node identification model takes into account the location of the road network nodes, the number of neighboring nodes, the influence of the neighboring nodes, and the path information between the nodes, and the key nodes in the network are effectively identified through the improvement of the gravitational model and the neighboring core model, and the model's validity is verified through the SIR contagion model. The online network interchange stations siting model considers the matching degree of residents' traffic demand, the interchange stations coverage to the city, the intensity of land use, and the interchange stations' duplicate coverage to the urban space as evaluation indices for the interchange stations; the objective of solving the interchange stations is the comprehensive optimality of the indexes; in order to avoid the problem of perturbation in the optimization and solution algorithm, this paper presents an improved genetic-simulated annealing optimization algorithm that introduces the concept of community, and avoids the problem of perturbation in the optimization and solution algorithms. The effectiveness of the model is verified by using examples of the Mitte-Center and Friedrichshain networks.

Emissions and Flame Stability Assessment of Hydrogen Addition and Air Dilution in a Micro Gas Turbine Combustor Using 0D/1D Modeling by Chemical Reactor Network

Abstract Hydrogen emerges as a promising fuel for clean and sustain- able electricity production when utilized in micro gas turbines (mGTs). However, some challenges linked to hydrogen usage must be overcome as its high reactivity can lead to increased NOx emissions and flame instabilities. Literature suggests that combustion air humidification and/or exhaust gas recirculation (EGR) may be methods to reduce this reactivity. However, these measures are limited due to the small operating range of the mGT combustor. In this context, a 20 kWth mGT combustor is investigated un- der various inlet conditions to assess the effect of EGR towards increased flame stability and emission control when operating under hydrogen-enriched methane firing, using a Chemical Re- action Network (CRN) model. The CRN modeling is a fast and low computational complexity tool to model combustion perfor- mance and emissions by representing complex reactive flow fields through idealized reactor models, leading to reduced computa- tional costs. This study examines partial load conditions, fuel compositions, and the effect of combustion air dilution through EGR. The CRN model of the combustion process is designed based on combustion zones extracted from the main flow fields with similar thermo-chemical states from CFD, while emission predictions are experimentally validated for different operating points. Predicted NOx and CO emissions by the proposed CRN model show an acceptable agreement with the experimental data at high power loads. However, its accuracy diminishes for loads below 70% due to the variability in model tuning parameters across different loads, whereas the

A causal network model to estimate the cardiotoxic effect of oncological treatments in young breast cancer survivors

A causal network model to estimate the cardiotoxic effect of oncological treatments in young breast cancer survivors

Predicting FRP Plate End Debonding with a Neural Network Model Enhanced by Modified Sparrow Search Algorithm

Predicting FRP Plate End Debonding with a Neural Network Model Enhanced by Modified Sparrow Search Algorithm

Sensory experience steers representational drift in mouse visual cortex

Representational drift—the gradual continuous change of neuronal representations—has been observed across many brain areas. It is unclear whether drift is caused by synaptic plasticity elicited by sensory experience, or by the intrinsic volatility of synapses. Here, using chronic two-photon calcium imaging in primary visual cortex of female mice, we find that the preferred stimulus orientation of individual neurons slowly drifts over the course of weeks. By using cylinder lens goggles to limit visual experience to a narrow range of orientations, we show that the direction of drift, but not its magnitude, is biased by the statistics of visual input. A network model suggests that drift of preferred orientation largely results from synaptic volatility, which under normal visual conditions is counteracted by experience-driven Hebbian mechanisms, stabilizing preferred orientation. Under deprivation conditions these Hebbian mechanisms enable adaptation. Thus, Hebbian synaptic plasticity steers drift to match the statistics of the environment.

Spiking neural network model of low-threshold mechanoreceptors system

Spiking neural network model of low-threshold mechanoreceptors system