Network Analysis Model Research Articles

A hybrid Network Analysis and Machine Learning model for Enhanced Financial Distress Prediction

A hybrid Network Analysis and Machine Learning model for Enhanced Financial Distress Prediction

Evaluation of the Influence of "Village Super League" and "Village BA" on the County Economy of Hunan Province Based on Network Analysis and XGBoost Model

Evaluation of the Influence of "Village Super League" and "Village BA" on the County Economy of Hunan Province Based on Network Analysis and XGBoost Model

Proposing A Conceptual Model for Asnafpreneur Success: Social Networks and Resource Dependence in Low-Income Regions

This paper proposes a conceptual model integrating Social Network Theory (SNT) and Resource Dependence Theory (RDT) to examine the business performance of Asnafpreneurs—entrepreneurs from marginalized communities supported by zakat—in low-income regions. While Asnafpreneurs frequently rely on social networks to access resources, they face significant challenges, including limited network diversity and heavy dependency on external actors like zakat institutions for financial and operational support. The model presented in this study aims to demonstrate how optimizing social networks can help reduce resource dependency and improve business outcomes, such as revenue growth, sustainability, and entrepreneurial autonomy. To guide future empirical validation, the paper suggests a mixed-methods approach, combining qualitative insights from in-depth interviews with quantitative measures, including social network analysis and structural equation modeling, to test the proposed relationships. This approach provides a roadmap for assessing the impact of social networks on business performance and understanding the mediating role of resource dependency. The findings contribute to the theoretical development of SNT and RDT by extending their application to marginalized entrepreneurship within a faith-based context. The paper also offers practical implications for policymakers and zakat institutions, providing strategies for designing more effective entrepreneurial support programs aimed at empowering Asnafpreneurs in low-income communities.

Network analysis and molecular modeling studies of pinocembrin a bioactive phytochemical of Dodonaea viscosa against Parkinson's disease.

Parkinson's disease, a neurodegenerative disorder, is quickly progressing and accounts for 15% of dementia cases. Parkinson's disease is the second most frequent form of neuronal degeneration after Alzheimer's, with an average age of 55 years for individuals exhibiting neuropsychiatric and physiological symptoms. Due to the effectiveness, low toxicity, and low side effects, bioactive compounds from plants have received increased attention recently as therapeutic drugs. In the current study, effective anti-neurodegenerative phytochemicals from Dodonaea viscosa were screened using in silico methods and have been proposed to be further investigated for the treatment of Parkinson's disease. The structures of twenty bioactive chemicals were screened and graph theoretical network analysis revealed alpha-synuclein as a potent therapeutic target. Based on docking scores, an effective bioactive molecule was selected, and its energy values, electrostatic potential surface and drug-like qualities were examined using molecular orbitals, pharmacokinetics and toxicity studies. Pinocembrin was found as a superior binder based on molecular docking as it demonstrated stronger binding with - 10.2kcal/mol. An investigation using Ramachandran plot validated the protein-ligand complex secondary structure's stability. Pinocembrin, a bioactive phytochemical from Dodonaea viscosa, may be a viable lead molecule that may be developed as a candidate medicine for anti-neurodegenerative therapy against Parkinson's disease.

Predicting hotspots for disease-causing single nucleotide variants using sequences-based coevolution, network analysis, and machine learning.

To enable personalized medicine, it is important yet highly challenging to accurately predict disease-causing mutations in target proteins at high throughput. Previous computational methods have been developed using evolutionary information in combination with various biochemical and structural features of protein residues to discriminate neutral vs. deleterious mutations. However, the power of these methods is often limited because they either assume known protein structures or treat residues independently without fully considering their interactions. To address the above limitations, we build upon recent progress in machine learning, network analysis, and protein language models, and develop a sequences-based variant site prediction workflow based on the protein residue contact networks: 1. We employ and integrate various methods of building protein residue networks using state-of-the-art coevolution analysis tools (RaptorX, DeepMetaPSICOV, and SPOT-Contact) powered by deep learning. 2. We use machine learning algorithms (Random Forest, Gradient Boosting, and Extreme Gradient Boosting) to optimally combine 20 network centrality scores to jointly predict key residues as hot spots for disease mutations. 3. Using a dataset of 107 proteins rich in disease mutations, we rigorously evaluate the network scores individually and collectively (via machine learning). This work supports a promising strategy of combining an ensemble of network scores based on different coevolution analysis methods (and optionally predictive scores from other methods) via machine learning to predict hotspot sites of disease mutations, which will inform downstream applications of disease diagnosis and targeted drug design.

A study on the analysis and understanding of art works based on graph neural networks

Abstract With the development of the Internet, many art works are uploaded for people to enjoy. To make it more convenient for viewers to understand and analyze the content of art works. In this paper, we use the GNN graph neural network analysis model to extract the art style, creation period, color expression and other features of the works through the work feature saliency attention module, and classify them using the similarity metric. At the same time, we put forward the GNNMMVisRe model to explore the works from the visual multimodality. The analysis results concluded that the Impressionist style of art works had its heyday in the 18th-19th centuries, when the number of works reached 11,674. The Baroque classicism style reached its heyday in the mid-18th century with 5,921 pieces. In terms of color use, low saturation and low luminance color palettes were selected, with average values of 81.3% and 83.4%, respectively. Using the GNN model to categorize, analyze and study the art works improves the understanding and appreciation of the works. It is also significant for the further development of art.

Temporal Dynamic Synchronous Functional Brain Network for Schizophrenia Classification and Lateralization Analysis.

Available evidence suggests that dynamic functional connectivity can capture time-varying abnormalities in brain activity in resting-state cerebral functional magnetic resonance imaging (rs-fMRI) data and has a natural advantage in uncovering mechanisms of abnormal brain activity in schizophrenia (SZ) patients. Hence, an advanced dynamic brain network analysis model called the temporal brain category graph convolutional network (Temporal-BCGCN) was employed. Firstly, a unique dynamic brain network analysis module, DSF-BrainNet, was designed to construct dynamic synchronization features. Subsequently, a revolutionary graph convolution method, TemporalConv, was proposed based on the synchronous temporal properties of features. Finally, the first modular test tool for abnormal hemispherical lateralization in deep learning based on rs-fMRI data, named CategoryPool, was proposed. This study was validated on COBRE and UCLA datasets and achieved 83.62% and 89.71% average accuracies, respectively, outperforming the baseline model and other state-of-the-art methods. The ablation results also demonstrate the advantages of TemporalConv over the traditional edge feature graph convolution approach and the improvement of CategoryPool over the classical graph pooling approach. Interestingly, this study showed that the lower-order perceptual system and higher-order network regions in the left hemisphere are more severely dysfunctional than in the right hemisphere in SZ, reaffirmings the importance of the left medial superior frontal gyrus in SZ. Our code was available at: https://github.com/swfen/Temporal-BCGCN.

Investigation of Dielectric Properties of Potato, Tomato and Onion Juices at Microwave Frequencies

The dielectric properties of fresh juices of Solanum tuberosum (potato), Solanum lycopersicum (tomato) and Allium cepa (onion) were measured in terms of the dielectric constant and dielectric loss factor, over a frequency range from 1 to 50 GHz and at temperatures ranging from 30 to 60 °C, by using the PNA network analyzer model E8364C and open-ended coaxial probe 85070E. A prediction of relaxation frequencies of molecules for the three different juices was done and their molecular behavior was then studied by plotting dielectric constant vs. frequency and dielectric loss factor vs. frequency curves in the frequency range from 1 to 50 GHz at four different temperatures from 30 to 60 °C. The dielectric properties of food products play a crucial role in determining their heating behavior when subjected to an electromagnetic field. Understanding these properties is essential for the development of microwave pasteurization and sterilization processes tailored to various food products.

A Network Intrusion Detection Method for Information Systems Using Federated Learning and Improved Transformer

A network intrusion detection method for information systems using federated learning and improved transformer is proposed to address the problems of long detection time and low security and accuracy when analyzing massive data in most existing intrusion detection methods. Firstly, a network intrusion detection system is constructed based on a federated learning framework, and the transformer model is used as its universal detection model. Then, the dataset is divided and an improved generative adversarial network is used for data augmentation to generate a new sample set to overcome the influence of minority class samples. At the same time, the new samples are input into the transformer local model for network attack type detection and analysis. Finally, the authors aggregate the detection results of each local model and input them into the Softmax classifier to obtain the final classification prediction results.

The effect of two facets of physicians’ environmental stress on patients’ compliance with COVID-19 guidelines: moderating roles of two types of ego network

Drawing upon the Conservation of Resources Theory, this study seeks to examine the association between two dimensions of environmental stress experienced by physicians and patients’ adherence to COVID–19 guidelines, within the context of a social network framework. A third—wave longitudinal study was employed to gather 439 valid data points in China. Social network analysis and structural equation model were used to test the conceptual model. The results reveal the pivotal role of physicians’ environmental stress related to their work and family contexts in influencing patients’ adherence to COVID–19 guidelines through the mediation of physicians’ information sharing. The ego networks of physicians, encompassing both advice–seeking and friendship ties, were observed to negatively moderate the relationship between stress and resource depletion. Broadly, our study shows the importance of understanding physicians’ stress caused by the working and family environments, as these factorsnot only impact the psychological well–being of physicians but also significantly affect patients’ compliance with COVID–19 guidelines. In addition, the work offers a framework for understanding the impact of the ego advice–seeking network and the ego friend network.

Multidimensional evolutionary analysis of China’s BIM technology policy based on quantitative mapping

ABSTRACT In the wave of the digital era, Building Information Modelling (BIM) has emerged as the primary tool for intelligent construction and digital transformation in the Architecture, Engineering, and Construction (AEC) industry. China's focus on policy guidance for the dissemination and application of BIM technology has promoted high-quality development and technological innovation in the AEC industry. However, few studies have described the multi-dimensional evolution picture of BIM technology policy in China, and lack of effective grasp and prediction of policy evolution law and future direction. An improved policy quantitative analysis framework is built using natural language processing (NLP) based on deep learning in this paper. The dynamic characteristics of 242 Chinese BIM technology policy documents from 1998 to 2022 in the PKULAW database are quantitatively analyzed by the framework, and the four-stage integrated evolution of Chinese BIM technology policy is quantitatively mapped in a visual form through network analysis and dynamic topic model (DTM). Furthermore, this research conducts a comprehensive discussion from three dimensions, providing valuable insights for policy makers and industry participants in China. The findings contribute to the deep integration of policy optimization and the construction industry, while also aiming to share China's experience in BIM technology governance with other countries and regions. Ultimately, this study seeks to support the global AEC industry in its digital transformation and development by offering valuable knowledge and perspectives.

A new network analysis model in anorexia nervosa patients based on self-reported eating disorder symptoms, psychological distress, and cognitive flexibility.

Cognitive flexibility and psychological distress, such as depression and anxiety, have been implicated in the aetiology of Anorexia Nervosa (AN). Despite the known associations between eating disorder (ED) symptoms, depression, anxiety, and cognitive flexibility, the specific pathways that connect these constructs are unclear. We therefore used network analysis to examine the relationship between these symptoms in an AN sample. One hundred and ninety-three treatment-seeking individuals diagnosed with AN (95.6% female, M = 26.89 [SD = 9.45] years old) completed self-report measures assessing depression, anxiety, cognitive flexibility, and ED symptoms. To determine each symptom's influence in the network, we calculated the expected influence. The two relationships with the greatest edges were those between (1) weight/shape concerns and eating/dietary restraint and (2) weight/shape concerns and psychological distress (a measure that combined depression and anxiety). Cognitive flexibility was not connected to weight/shape concerns but had negative partial associations with eating concerns/dietary restraint and psychological distress. There was also a slight, non-zero connection between eating concerns/dietary restraint and psychological distress. The findings underscore the importance of weight/shape, eating/dietary concerns, and psychological distress in the AN network and suggest that addressing cognitive flexibility may be a useful target for eating concerns/dietary restraint and psychological distress. Future studies assessing the longitudinal course of psychopathology within the AN network structure may help in identifying whether specific symptoms function as risk factors or maintaining factors for this co-occurrence.

Risk spillover between carbon markets and stock markets from a progressive perspective: Measurements, spillover networks, and driving factors

Prior studies have established the existence of risk correlation and spillover effects between carbon markets and stock markets. However, these research findings have mostly been at the whole market-level and are limited in their applicability to sectoral-level and enterprise-level decision-making, especially for emerging economies like China. This paper adopts an innovative and progressive perspective to bridge this existing gap by leveraging the Diebold-Yilmaz spillover index method, social network analysis and exponential random graph model to comprehensively evaluate the spillover level, network structure, and driving factors between the China's carbon market and the stock markets across various sectors. It is found that: (1) the China's carbon market and stock market as a whole display significant, unbalanced, and extreme-event-sensitive spillover effects, with the carbon market primarily serving as a net receiver of information; (2) the spillover effects exhibit noteworthy heterogeneities across diverse sectoral stock markets, and we discovered that the carbon market plays the role of a “main benefit ” in the network; and (3) within the selected four representative sectors, environmental information disclosure indicator, enterprise green transformation indicator and selected financial indicators all have the potential to significantly influence the structure of inter-enterprise spillover networks to varying degrees. Overall, the findings of our study hold significant practical implications for policymakers and investors involved in the carbon market, while providing new insights into addressing real issues related to sectoral coverage in the carbon market.

The unique role of smartphone addiction and related factors among university students: a model based on cross-sectional and cross-lagged network analyses

Smartphone addiction is a global problem affecting university students. Previous studies have explored smartphone addiction and related factors using latent variables. In contrast, this study examines the role of smartphone addiction and related factors among university students using a cross-sectional and cross-lagged panel network analysis model at the level of manifest variables. A questionnaire method was used to investigate smartphone addiction and related factors twice with nearly six-month intervals among 1564 first-year university students (M = 19.14, SD = 0.66). The study found that procrastination behavior, academic burnout, self-control, fear of missing out, social anxiety, and self-esteem directly influenced smartphone addiction. Additionally, smartphone addiction predicted the level of self-control, academic burnout, social anxiety, and perceived social support among university students. Self-control exhibited the strongest predictive relationship with smartphone addiction. Overall, self-control, self-esteem, perceived social support, and academic burnout were identified as key factors influencing smartphone addiction among university students. Developing prevention and intervention programs that target these core influencing factors would be more cost-effective.

Combining with volatilomic profiling and chemometrics to explore the volatile characteristics in five different dried Zanthoxylum bungeanum maxim

Owing to the short picking period of the fresh Zanthoxylum bungeanum, the postharvest drying has become an essential operation before the storage and transportation of Z. bungeanum. To explore the effects of drying methods on volatile characteristics, the volatilomic profiling of five different dried Z. bungeanum was investigated by E-nose, HS-SPME-GC/MS, GC-IMS in combination with chemometrics. The results indicated that W1W, W2W and W5S sensors within E-nose analysis showed the strongest responses in both fresh and dried Z. bungeanum. According to the identification of volatile organic compounds (VOCs), terpenes, esters and alcohols played the major roles in the volatile formation of the fresh and dried Z. bungeanum. The samples derived from hot air drying showed the relatively similar features with the fresh sample based on the relative abundances of these major VOCs. According to the results of multiple factor analysis (MFA), GC-IMS showed the strongest ability in distinguishing the fresh and different dried samples. Compared with the high levels of terpenes in fresh group, the significant increasement of terpene alcohols and terpene esters from the degradation and transformation of bound terpenoids was the main characteristics of all dried Z. bungeanum. Using the GC-IMS datasets, a weighted correlation network analysis (WCNA) model was constructed to clarify the VOC characteristics in all detetected samples. Thereinto, 6 significantly correlated modules were identified in fresh and five different dried samples. Additionally, a total of 23 hub VOCs can be recognized as the potential biomarkers for better distinguishing the fresh and five different dried Z. bungeanum.

A new comprehensive quantitative index for the assessment of essential amino acid quality in beef using Vis-NIR hyperspectral imaging combined with LSTM

The quality of beef is usually predicted by measuring a single index rather than a comprehensive index. To precisely determine the essential amino acid (EAA) contents in 360 beef samples, the feasibility of optimized spectral detection techniques based on the comprehensive EAA index (CEI) and comprehensive weight index (CWI) constructed by factor analysis was explored. Two-dimensional correlation spectroscopy (2D-COS) was used to analyse the mechanisms of spectral peak shifts in complex disturbance systems with CEI and CWI contents, and 15 sensitive feature variables were extracted to establish a quantitative analysis model of a long short-term memory network (LSTM). The results indicated that 2D-COS had good predictive performance in both CEI-LSTM (R2P of 0.9095 and RPD of 2.76) and CWI-LSTM (R2P of 0.8449 and RPD of 2.45), which reduced data information by 88%. This indicates that utilizing 2D-COS can eliminate collinearity and redundant information among variables while achieving data dimensionality reduction and simplification of calibration models. Furthermore, a spatial distribution map of the comprehensive EAA content was generated by combining the optimal prediction model. This study demonstrated that the comprehensive index method furnishes a new approach to rapidly evaluate EAA content.

Identification of M5c regulator-medicated methylation modification patterns for prognosis and immune microenvironment in glioma.

Glioma is a common intracranial tumor and is generally associated with poor prognosis. Recently, numerous studies illustrated the importance of 5-methylcytosine (m5C) RNA modification to tumorigenesis. However, the prognostic value and immune correlation of m5C in glioma remain unclear. We obtained RNA expression and clinical information from The Cancer Genome Atlas (TCGA) and The Chinese Glioma Genome Atlas (CGGA) datasets to analyze. Nonnegative matrix factorization (NMF) was used to classify patients into two subgroups and compare these patients in survival and clinicopathological characteristics. CIBERSORT and single-sample gene-set algorithm (ssGSEA) methods were used to investigate the relationship between m5C and the immune environment. The Weighted correlation network analysis (WGCNA) and univariate Cox proportional hazard model (CoxPH) were used to construct a m5C-related signature. Most of m5C RNA methylation regulators presented differential expression and prognostic values. There were obvious relationships between immune infiltration cells and m5C regulators, especially NSUN7. In the m5C-related module from WGCNA, we found SEPT3, CHI3L1, PLBD1, PHYHIPL, SAMD8, RAP1B, B3GNT5, RER1, PTPN7, SLC39A1, and MXI1 were prognostic factors for glioma, and they were used to construct the signature. The great significance of m5C-related signature in predicting the survival of patients with glioma was confirmed in the validation sets and CGGA cohort.

Dimethyl fumarate treatment in relapsing remitting MS changes the inflammatory CSF protein profile by a prominent decrease in T-helper 1 immunity

BackgroundDimethyl fumarate (DMF) is a common treatment for multiple sclerosis (MS), but its mechanisms of action are not fully understood. Targeted proteomics offers insights into effects of DMF and biomarkers for treatment responses. ObjectivesTo assess influence of DMF on inflammation- and neuro-associated proteins in plasma and cerebrospinal fluid (CSF) in MS and to reveal biomarkers for predicting treatment responses. MethodsUsing the high-sensitivity and high-specificity method of proximity extension assay (PEA), we measured 182 inflammation- and neuro-associated proteins in paired plasma (n = 28) and CSF (n = 12) samples before and after one year of DMF treatment. Disease activity was evaluated through clinical examination and MRI. Statistical tests, network analysis, and regression models were used. ResultsSeveral proteins including T-helper 1 (Th1)-associated proteins (CXCL10, CXCL11, granzyme A, IL-12p70, lymphotoxin-alpha) were consistently decreased in CSF, while IL-7 was increased after one year of treatment. The changes in plasma protein levels did not follow the same pattern as in CSF. Logistic regression models identified potential biomarker candidates (including plexins and neurotrophins) for prediction of treatment response. ConclusionsDMF treatment induced prominent changes in CSF proteins, consistently reducing Th1-associated pro-inflammatory proteins. Neurodegeneration-related CSF proteins were able to predict treatment response. Protein biomarkers hold promise for personalized medicine.

Probabilistic risk assessment of fire and explosion of onboard high-pressure hydrogen system

Hydrogen is regarded as a new energy carrier and has broad application prospects. As the leading field of hydrogen energy application, hydrogen fuel cell vehicles (HFCVs) are highly valued in the world. However, due to the flammable and explosive characteristics of hydrogen, its safety issues have attracted much attention, and it is increasingly necessary to understand the possible accident risks of the hydrogen system of HFCVs. In this study, the quantitative risk assessment (QRA) of hydrogen system of HFCV was carried out, and the Bayesian network (BN) analysis model of fire and explosion risk of onboard high-pressure hydrogen system was established. Taking a hydrogen system as an example, BN parameters are quantified based on three quantitative analysis methods: general data, logical judgment and computational simulation, and then the fire and explosion risk and personnel loss degree of the general onboard high-pressure hydrogen system are determined. The results show that when hydrogen release occurs, the probability of explosion is 6.79E-05, the probability of jet fire is 1.53E-04, and the probability of fireball is 5.38E-08. In the case of major damage, there is an 87.15 % probability that an explosion occurred; the reliability of safety protection devices is very important in the process of hydrogen use. It is suggested to optimize the structural layout design of safety components and improve the reliability of protection devices to promote the safe development of hydrogen energy industry such as HFCVs.

Network Analysis of Posttraumatic Growth Dimensions: A Cross-Sectional Study in People Who Experienced the Death of a Loved One from COVID-19 in 16 Latin American Countries

The present study aimed to apply a network analysis model to provide an exploratory empirical conceptualization of dynamic networks of posttraumatic growth (PTG) symptoms in 7,434 people who experienced the death of a loved one from COVID-19 in 16 Latin American countries. The Post-Traumatic Growth Inventory: Short Form of Eight Items was used. A non-regularized network with partial correlation coefficients was estimated through the ggmModSelect algorithm. The network architecture was analyzed for each country through its local properties and global properties. The results indicated that the networks differed significantly between countries. The core dimensions in the networks were relating to others, personal strength, and life value and opportunities, which were more related dimensions that reinforce the emergence of PTG in all countries. The findings may be useful to motivate researchers and mental health professionals to consider the importance of the individual dimensions of PTG in groups of people who experienced the death of a loved one from COVID-19 in 16 Latin American countries, as well as their interrelationships.