Graphical Models Research Articles

Modeling and simulation of liquid metal battery (LMB) materials development using statistical package for social sciences (S.P.S.S.)

There are a number of input parameters that are considered in relation to the stimulatory possibility of constructing a Liquid Metal Battery (LMB). This paper talks about the modeling approach possible for use in LMB research work. Equivalent Circuit Modeling (E.C.M.) is the most common method used to analyze input data or parameters. In analyzing some of the basic elements, such as electrical capacitance, electrical resistance, open circuit voltage, terminal voltage, temperature, time response, time constants, State of Charge (SOC), etc. The cell impedance could be calculated by predicting the system elements that would play key roles in the determination of the parameter identification method for the battery’s Equivalent Circuit Model. Secondly, each element in the model has a known behaviour, which mainly depends on the element type and the values of the parameters that characterize that element. In EIS software, the Graphical Model Editor could be used to build an equivalent circuit model, or the befitting model could be carefully and properly selected. Thirdly, in fitting the Equivalent Circuit Model (ECM) to the initial data or parameters, one must take note that the parameters are strongly dependent on temperature, heat, and losses. Such are: the Open Circuit Voltage (OCV), which is strongly dependent on the temperature, the loss processes depend on temperature; losses produced by the loss processes are dissipated as heat energy; the heat generated or consumed by the electrochemical reactions during normal operation has a defined temperature; and a system with a defined temperature window with safe, stable, and efficient operation is achievable. At the end of this research, the simulation of Lithium (Li) and Cadmium (Cd) was found to be in the proportion of 67:33, which is used to determine the strength of the reactivity of the metals. It can be informed in this article that Bi and Li chemical compositions of the metals are equal to 49 for Li and 51 for Bismuth, which makes the overall reactivity very high, which could be used for LMB development.

Abstract WMP99: A Novel Network Analysis of County- and Individual-Level Factors Associated With Functional Outcomes After Stroke

Background: Social circumstances may contribute to health disparities in patients with stroke through indirect and interdependent mechanisms that are not well understood. Objective: This study used complexity science methodology to examine the interplay between county- and individual-level social and clinical factors influencing stroke functional outcomes. Methods: As part of the Get With the Guidelines-Stroke (GWTG-S) Data Challenge, the GWTG-S Registry was merged with county-level Institute for Health Metrics and Evaluation data. Patients diagnosed with stroke (ischemic, subarachnoid, or intracerebral hemorrhage) were included. Multilayer networks were constructed by estimating mixed graphical models of 32 nodes across four layers, including social (county- and patient-level) and clinical (comorbidities and encounter) factors. Social determinant networks were estimated for patients with less favorable (ie., functional dependence or death, modified Rankin Score [mRS] 3-6) versus favorable outcomes (ie., functional independence, mRS 0-2). Network structure and node centrality were compared between mRS groups using bootstrap permutation analyses. Hub nodes were defined by betweenness centrality and facilitate effects across the network. Results: From 2013-2019, 990 721 (62.3%) stroke patients had mRS of 3-6 and 597 477 (37.6%) had mRS 0-2 at discharge. Compared to the mRS 0-2 group, the mRS 3-6 group’s social determinants network had greater global connectivity (p<0.001), and homelessness (p<0.001) and Black race (p<0.001) were hub nodes. Unique to the mRS 3-6 group, younger patients were more likely to identify as homeless (p=0.031), uninsured (p=0.001), and live in a county with lower per capita income (p<0.001). Conclusions: The effects of social determinants were significantly greater in patients with less favorable functional outcomes. Homelessness and race play critical roles in mediating the impacts of county- and individual-level social determinants on downstream disparities in stroke outcomes. Particularly among younger patients, housing, insurance status, and income may serve as a critical leverage point for county-level interventions to improve function after stroke.

Profiles associated with significant hepatic fibrosis consisting of alanine aminotransferase >30U/L, exercise habits, and metabolic dysfunction-associated steatotic liver disease.

In patients with steatotic liver disease (SLD), significant hepatic fibrosis is a prognostic factor with various etiologies, including inflammation and metabolic dysfunction. This study aimed to investigate independent factors and profiles associated with significant hepatic fibrosis, including alanine aminotransferase (ALT) levels >30U/L and metabolic dysfunction-associated SLD (MASLD), in health check-up examinees. This single-center, retrospective, observational cohort study enrolled 1378 consecutive health checkup examinees from April 2018 to September 2022. Shear wave elastography (SWE) was performed during a routine ultrasound examination, and patients with liver stiffness ≥6.60kPa were defined as having significant hepatic fibrosis. Patients were classified into nonsignificant hepatic fibrosis (n=1220) or a significant hepatic fibrosis (n=158) group according to this definition. In multivariate analysis, the independent factor for significant hepatic fibrosis was aging (≥65years; OR 9.637, 95% CI 6.704-13.852, p<0.0001). According to decision tree analysis, the initial classifier was aging (≥65years). After aging, an ALT level >30U/L was the second relevant factor for significant hepatic fibrosis, regardless of age. An undirected graphical model showed that an ALT level of >30U/L was directly associated with significant hepatic fibrosis. In patients aged ≥65years with an ALT level >30U/L, significant hepatic fibrosis was observed in 52% of the patients. Meanwhile, in patients aged ≥65years with an ALT level ≤30U/L, MASLD was the third classifier, with significant hepatic fibrosis observed in 38% of patients. ALT levels >30U/L and MASLD may be involved in the pathogenesis of significant hepatic fibrosis in patients aged ≥65years.

Clarifying directional dependence among measures of early auditory processing and cognition in schizophrenia: leveraging Gaussian graphical models and Bayesian networks.

Research using latent variable models demonstrates that pre-attentive measures of early auditory processing (EAP) and cognition may initiate a cascading effect on daily functioning in schizophrenia. However, such models fail to account for relationships among individual measures of cognition and EAP, thereby limiting their utility. Hence, EAP and cognition may function as complementary and interacting measures of brain function rather than independent stages of information processing. Here, we apply a data-driven approach to identifying directional relationships among neurophysiologic and cognitive variables. Using data from the Consortium on the Genetics of Schizophrenia 2, we estimated Gaussian Graphical Models and Bayesian networks to examine undirected and directed connections between measures of EAP, including mismatch negativity and P3a, and cognition in 663 outpatients with schizophrenia and 630 control participants. Chain structures emerged among EAP and attention/vigilance measures in schizophrenia and control groups. Concerning differences between the groups, object memory was an influential variable in schizophrenia upon which other cognitive domains depended, and working memory was an influential variable in controls. Measures of EAP and attention/vigilance are conditionally independent of other cognitive domains that were used in this study. Findings also revealed additional causal assumptions among measures of cognition that could help guide statistical control and ultimately help identify early-stage targets or surrogate endpoints in schizophrenia.

A graphical method for causal program attribution in theory-based evaluation

We describe a simple yet rigorous graphical method for eliminating bias in theory-based program evaluation. The method is an application to social and international development program evaluation of the graphical causal models used to test medical treatments. We implement a graphical causal model for the World Bank’s well-known Bangladesh Integrated Nutrition Project. We show how to construct the graphical causal model to represent program theory in context in explicitly causal terms. We then show how to visually inspect the graphical causal model to distinguish causal from non-causal associations between variables in evaluation data. Finally, we show how to select a set of adjustment variables to neutralize non-causal associations, eliminating bias in all forms of causal inference—qualitative and quantitative, linear and non-linear.

Integrative pathway analysis with gene expression, miRNA, methylation and copy number variation for breast cancer subtypes.

Developments in biotechnologies enable multi-platform data collection for functional genomic units apart from the gene. Profiling of non-coding microRNAs (miRNAs) is a valuable tool for understanding the molecular profile of the cell, both for canonical functions and malignant behavior due to complex diseases. We propose a graphical mixed-effects statistical model incorporating miRNA-gene target relationships. We implement an integrative pathway analysis that leverages measurements of miRNA activity for joint analysis with multimodal observations of gene activity including gene expression, methylation, and copy number variation. We apply our analysis to a breast cancer dataset, and consider differential activity in signaling pathways across breast tumor subtypes. We offer discussion of specific signaling pathways and the effect of miRNA integration, as well as publish an interactive data visualization to give public access to the results of our analysis.

Combination of autoregressive graphical models and time series bootstrap methods for risk management in marine insurance

In this paper a methodology to assess risk by forecasting the trend of marine losses at a global scale is presented. The proposed procedure, which can be used to continuously update an insurance company’s costing model, identifies the most relevant risk indicators through Probabilistic Graphical Models (PGMs). The use of PGMs makes the variable selection more understandable since they provide a clear interface to interpret the model and perform predictions. Furthermore, this procedure can be used to verify independence relationships, validate the dataset and identify unexpected links among the considered variables. The robustness of estimates, crucial for risk assessment in the insurance context, is dealt with bootstrap.

Personality traits and depressive symptoms among Chinese older people: A network approach

BackgroundThe present study aims to investigate the network structure of depressive symptoms, the interrelationships between individual personality traits and depressive symptoms, and gender differences among Chinese older people aged 60 and above. MethodWe performed network analyses with a regularized Graphical Gaussian Model and a case-dropping bootstrap approach. A sample of 4876 older Chinese people aged 60+ was included in the analyses. We investigated the central symptoms in the depression network and the bridge nodes that connect personality facets and depressive symptoms. Gender differences were investigated by testing the global strength, network invariance, and edge weights. ResultsSadness and depressed mood were the most central depressive symptoms, while somatic symptoms such as restless sleep were the least central. Neurotic facets, particularly “worry a lot” and “get nervous easily”, played significant bridging roles in the web of personality traits and depressive symptoms. Gender differences were observed in three edges among different personality traits (rude-worried, original-worried, and forgiving-nervous). LimitationThe study adopts a cross-sectional dataset, and therefore, cannot track the network changes over time or conclude a causal relationship. DiscussionThe study calls for more focus and prioritization on sadness, depressed mood and neurotic traits in the identification of depression among older Chinese people. Future researchers and practitioners should better understand of older Chinese adults' worry and nervousness to develop appropriate practices and policies.

Student teachers’ opportunities to learn through observation, own practice and feedback on the practice while in field practice placements: a graphical model approach

Field practice placement is a crucial part of teacher education, as it affords a real-life context, where teacher and teacher-related skills can be enacted and trained. The present study examined the associations between student teacher opportunities to learn through observation, own practice and the receiving of feedback of said practice, while in field practice placements through a teacher education programme. Chain graph models were used to analyse data from 560 Danish student teachers who had just completed field practice at one of three levels. Results showed that opportunities to learn through observation of fellow students and other teachers was negatively associated with level of field practice, and thus was reported less and less the further along students were in the programme, while opportunities to learn through own practice was positively associated with level of field practice. Opportunities to learn through receiving feedback on own practice was associated with level of field practice only via opportunities to learn through own practice. Results did not reveal gender or age-wise inequity in the opportunities to learn afforded in the field practice. Teacher education programmes could benefit from placing additional focus on opportunities to learn through observation in the later field practice placements.

Exploring the complexity of commonly held attitudes and beliefs of low back pain-a network analysis.

The current study used a network analysis approach to explore the complexity of attitudes and beliefs held in people with and without low back pain (LBP). The study aimed to (1) quantify the adjusted associations between individual items of the Back Pain Attitudes Questionnaire (Back-PAQ), and (2) identify the items with the strongest connectivity within the network. This is a secondary data analysis of a previously published survey using the Back-PAQ (n = 602). A nonparametric Spearman's rank correlation matrix was used as input to the network analysis. We estimated an unregularised graphical Gaussian model (GGM). Edges were added or removed in a stepwise manner until the extended Bayesian information criterion (EBIC) did not improve. We assessed three measures of centrality measures of betweenness, closeness, and strength. The two pairwise associations with the greatest magnitude of correlation were between Q30-Q31 [0.54 (95% CI 0.44 to 0.60)] and Q15-Q16 [0.52 (95% CI 0.43 to 0.61)]. These two relationships related to the association between items exploring the influence of attentional focus and expectations (Q30-Q31), and feelings and stress (Q15-Q16). The three items with the greatest average centrality values, were Q22, Q25, and Q10. These items reflect beliefs about damaging the back, exercise, and activity avoidance, respectively. Beliefs about back damage, exercise, and activity avoidance are factors most connected to all other beliefs within the network. These three factors may represent candidate targets that clinicians can focus their counseling efforts on to manage unhelpful attitudes and beliefs in people experiencing LBP.

THE ROLE OF UKRAINE IN THE FORMATION OF WORLD FOOD SECURITY

The authors considered the role of Ukraine in the formation of world food security. It was established that the export of agricultural products is a priority component of the national economy. It is studied how the full-scale invasion of the Russian Federation on the territory of sovereign independent Ukraine affects the export of Ukrainian food. The state of Ukrainian food exports based on the results of the 2021/2022 marketing year and 2022/2023 marketing year was considered. The specifics of Ukrainian export of food and agricultural goods are analyzed. The export potential of Ukraine has been proven in the conditions of protracted military aggression and large-scale Russian missile attacks on the entire territory of Ukraine, which have continued continuously for more than one year. The authors considered current operational economic indicators and analyzed the dynamics of their decline compared to pre-war data. The reasons for the decrease in grain supplies have been established. Forecasts of various experts regarding exports for the near future were studied. The authors presented relevant data and graphical models for a more accessible understanding of the problem. The general impact of the Russian-Ukrainian war on world food security is substantiated. In the work, the authors emphasized the importance of Ukrainian food supplies at the international level.

Graphical representation of data prediction potential: correlation graphs and correlation chains

The correlation of the set of attributes is a crucial statistical value for the measuring of prediction potential present in a dataset. The correlation coefficient, which measures the correlation between the values of two attributes, can be used in order to measure the prediction potential between two-element subsets of a dataset containing a high number of attributes. In this way two common summary visualizations of prediction potential in datasets are formed—correlation matrices and correlation heatmaps. Both of these visualizations are focused on the presentation of correlation between pair of attributes but not much more regarding the context of correlations in the dataset. The main objective of this article is the design and implementation of graphical models usable in a visual representation of data prediction potential—correlation graphs and correlation chains—which emphasize the pseudo-transitivity of prediction potential in a dataset.

Generation of probabilistic synthetic data for serious games: A case study on cyberbullying

Synthetic data generation has been a growing area of research in recent years. However, its potential applications in serious games have yet to be thoroughly explored. Advances in this field could anticipate data modeling and analysis, as well as speed up the development process. To fill this gap in the literature, we propose a simulator architecture for generating probabilistic synthetic data for decision-based serious games. This architecture is designed to be versatile and modular so that it can be used by other researchers on similar problems (e.g., multiple choice exams, political surveys, any type of questionnaire). To simulate the interaction of synthetic players with the game, we use a cognitive testing model based on the Item Response Theory framework. We also show how probabilistic graphical models (in particular, Bayesian networks) can introduce expert knowledge and external data into the simulation. Finally, we apply the proposed architecture and methods in the case of a serious game focused on cyberbullying. We perform Bayesian inference experiments using a hierarchical model to demonstrate the identifiability and robustness of the generated data.

Integrative analysis of noncoding mutations identifies the druggable genome in preterm birth.

Preterm birth affects ~10% of pregnancies in the US. Despite familial associations, identifying at-risk genetic loci has been challenging. We built deep learning and graphical models to score mutational effects at base resolution via integrating the pregnant myometrial epigenome and large-scale patient genomes with spontaneous preterm birth (sPTB) from European and African American cohorts. We uncovered previously unidentified sPTB genes that are involved in myometrial muscle relaxation and inflammatory responses and that are regulated by the progesterone receptor near labor onset. We studied genomic variants in these genes in our recruited pregnant women administered progestin prophylaxis. We observed that mutation burden in these genes was predictive of responses to progestin treatment for preterm birth. To advance therapeutic development, we screened ~4000 compounds, identified candidate molecules that affect our identified genes, and experimentally validated their therapeutic effects on regulating labor. Together, our integrative approach revealed the druggable genome in preterm birth and provided a generalizable framework for studying complex diseases.

Copula Graphical Models for Heterogeneous Mixed Data

This article proposes a graphical model that handles mixed-type, multi-group data. The motivation for such a model originates from real-world observational data, which often contain groups of samples obtained under heterogeneous conditions in space and time, potentially resulting in differences in network structure among groups. Therefore, the iid assumption is unrealistic, and fitting a single graphical model on all data results in a network that does not accurately represent the between group differences. In addition, real-world observational data is typically of mixed discrete-and-continuous type, violating the Gaussian assumption that is typical of graphical models, which leads to the model being unable to adequately recover the underlying graph structure. Both these problems are solved by fitting a different graph for each group, applying the fused group penalty to fuse similar graphs together and by treating the observed data as transformed latent Gaussian data, respectively. The proposed model outperforms related models on learning partial correlations in a simulation study. Finally, the proposed model is applied on real on-farm maize yield data, showcasing the added value of the proposed method in generating new production-ecological hypotheses. An R package containing the proposed methodology can be found on https://CRAN.R-project.org/package=heteromixgm. Supplementary materials for this article are available online.

Identification of Criteria for Enabling the Adoption of Sustainable Maintenance Practice: An Umbrella Review

The evolution from traditional industrial maintenance to sustainable maintenance (SM) is pivotal within an existing industrial ecosystem. This study, utilising an umbrella review (UR), critically examines this transition, highlighting its increased importance in maintenance decision-making (MDM). Using a sample (n = 20) of reviews, we synthesised meta-, methodological-, and content-based evidence and performed bibliometric, thematic and statistical analyses. For the bibliometric and thematic/conceptual analyses, we used the R bibliometrix package. The results show that the early research focuses mainly on theoretical aspects, while recent studies examine the practical implications. Also, comprehensive studies evaluating the benefits of implementing environmental and social aspects within MDM are still lacking. For that reason, we switched the attention to content-based data, from which we identified 43 distinct criteria discussed. For the analysis of criteria, the Bayesian Network Analysis with Gaussian Copula Graphical Model (BNA-GCGM) method was used. Although the evidence shows that environmental pollution, energy consumption and health and safety of workers are the most discussed criteria, the BNA-GCGM suggests that labour costs, resource consumption, employee satisfaction and energy consumption, among others, are the most influential criteria in the network analysis. Interestingly, after distinguishing studies into pre- and post-2021 research, the results show that pre-2021 research is primarily focused on economic and technical factors, reflecting a profit-oriented approach. The post-2021 analysis suggests a discernible shift towards more balanced considerations by incorporating social and environmental factors, suggesting a more socially responsible approach. Finally, while SM is gaining momentum, further empirical and practical research are required to demonstrate the advantages that SM offers in the light of the upcoming Industry 5.0.

DNBP: Differentiable Nonparametric Belief Propagation

We present a differentiable approach to learn the probabilistic factors used for inference by a nonparametric belief propagation algorithm. Existing nonparametric belief propagation methods rely on domain-specific features encoded in the probabilistic factors of a graphical model. In this work, we replace each crafted factor with a differentiable neural network, enabling the factors to be learned using an efficient optimization routine from labeled data. By combining differentiable neural networks with an efficient belief propagation algorithm, our method learns to maintain a set of marginal posterior samples using end-to-end training. We evaluate our differentiable nonparametric belief propagation (DNBP) method on a set of articulated pose tracking tasks and compare performance with learned baselines. Results from these experiments demonstrate the effectiveness of using learned factors for tracking and suggest the practical advantage over hand-crafted approaches. The project webpage is available at: https://progress.eecs.umich.edu/projects/dnbp/ .

Reconstructing Disease Histories in Huge Discrete State Spaces

AbstractMany progressive diseases develop unnoticed and insidiously at the beginning. This leads to an observational gap, since the first data on the disease can only be obtained after diagnosis. Mutual Hazard Networks address this gap by reconstructing latent disease dynamics. They model the disease as a Markov chain on the space of all possible combinations of progression events. This space can be huge: Given a set of $$n\ge 266$$ n ≥ 266 events, its size exceeds the number of atoms in the universe. Mutual Hazard Networks combine time-to-event modeling with generalized probabilistic graphical models, regularization, and modern numerical tensor formats to enable efficient calculations in large state spaces using compressed data formats. Here we review Mutual Hazard Networks and put them in the context of machine learning theory. We describe how the Mutual Hazard assumption leads to a compact parameterization of the models and show how modern tensor formats allow for efficient computations in large state spaces. Finally, we show how Mutual Hazard Networks reconstruct the most likely history of glioblastomas.

Identifying the latent relationships between factors associated with traffic crashes through graphical models

Traffic safety field has been oriented toward finding the relationships between crash outcomes and predictor variables to understand crash phenomena and/or predict future crashes. In the literature, the main framework established for this purpose is based on constructing a modelling equation in which crash outcome (e.g., frequencies) is examined in relation to explanatory variables chosen based on the problem at hand. Despite the importance and success of this approach, there are two issues that are generally not discussed: 1) the latent relationships between factors associated with crashes are oftentimes not the focus of analysis or not observed; and 2) there are not many tools to make informed decisions on which variables might have an impact on the crash outcome and should be included in a safety model, particularly when observations are limited. To address these issues, this paper proposes the use of graphical models, namely a Markov random field (MRF) modelling, Bayesian network modelling, and a graphical XGBoost approach, to disclose relationship topologies of explanatory variables leading to fatal and incapacitating injury pedestrian crashes. The application of graph learning models in traffic safety has a high potential because they are not only useful to understand the mechanism behind the crash occurrence but also can assist in devising accurate and reliable prevention measures by identifying the true variable structure and essential factors jointly acting towards crash occurrence, similar to a pathological examination.

Chain event graphs for assessing activity-level propositions in forensic science in relation to drug traces on banknotes

Abstract Graphical models can be used to compare support given by evidence to propositions put forward by competing parties during court proceedings. Such models can also be used to evaluate support for activity-level propositions, that is, propositions referring to the nature of activities associated with evidence and how this evidence came to be at a crime scene. Graphical methods can be used to show different scenarios that might explain the evidence in a case and to distinguish between evidence requiring evaluation by a jury and quantifiable evidence from the crime scene. Such visual representations can be helpful for forensic practitioners, police, and lawyers who may need to assess the value that different pieces of evidence make to their arguments. In this article, we demonstrate how chain event graphs (CEGs) can be applied to a drug trafficking case. We show how different evidence (i.e. expert judgement and data from a crime scene) can be combined using a CEG and show how the hierarchical model deriving from the graph can be used to evaluate the degree of support for different activity-level propositions in the case. We also develop a modification of the standard CEG to simplify its use in forensic applications.