Regulations of cosmetic ingredients and products have been the most advanced in embracing new approach methodologies (NAMs). Consequently, the cosmetic industry has assumed a forerunner role in the development and implementation of animal-free next-generation risk assessment (NGRA) that incorporates defined approaches (DAs) to assess the skin sensitization potency of ingredients. A Bayesian network DA predicting four potency categories (SkinSens-BN) was constructed against reference Local Lymph Node Assay data for a total of 297 substances, achieving a predictive performance similar to that of other DAs. With the aim of optimally informing risk assessment with a continuous point of departure (PoD), a weighted sum of the SkinSens-BN probabilities for four potency classes (non-, weak, moderate, and strong/extreme sensitizer) was calculated, using fixed weights based on associated LLNA EC3-values. The approach was promising, e.g., the derived PoDs for substances classified as non-sensitizers did not overlap with any others and 77% of PoDs were similar or more conservative than LLNA EC3. In addition, the predictions were assigned a level of confidence based on the probabilities to inform the evaluation of uncertainty in an NGRA context. In conclusion, the PoD derivation approach can substantially contribute to reliable skin sensitization NGRAs.

This study incorporated deep learning for periodontal disease detection into a Bayesian network (BN) clinical decision support model for comprehensive periodontal care. BN structure and probabilities were based on clinical data and Faster R-CNN-detected radiographic images. Receiver operating characteristic curve analysis confirmed the model\'s high accuracy in treatment plan recommendations.

Traditionally, firearm and toolmark examiners manually evaluate the similarity of features on two bullets using comparison microscopy. Advances in microscopy have made it possible to collect 3D topographic data, and several automated comparison algorithms have been introduced for the comparison of bullet striae using these data. In this study, open-source approaches for cross-correlation, congruent matching profile segments, consecutive matching striations, and a random forest model were evaluated. A statistical characterization of these automated approaches was performed using four datasets of consecutively manufactured firearms to provide a challenging comparison scenario. Each automated approach was applied to all samples in a pairwise fashion, and classification performance was compared. Based on these findings, a Bayesian network was empirically learned and constructed to leverage the strengths of each individual approach, model the relationship between the automated results, and combine them into a posterior probability for the given comparison. The network was evaluated similarly to the automated approaches, and the results were compared. The developed Bayesian network classified 99.6% of the samples correctly, and the resultant probability distributions were significantly separated more so than the automated approaches when used in isolation.

Considering activity level propositions in the evaluation of forensic biology findings is becoming more common place. There are increasing numbers of publications demonstrating different transfer mechanisms that can occur under a variety of circumstances. Some of these publications have shown the possibility of DNA transfer from site to site on an exhibit, for instance as a result of packaging and transport. If such a possibility exists, and the case circumstances are such that the area on an exhibit where DNA is present or absent is an observation that is an important diagnostic characteristic given the propositions, then site to site transfer should be taken into account during the evaluation of observations. In this work we demonstrate the ways in which site to site transfer can be built into Bayesian networks when carrying out activity level evaluations of forensic biology findings. We explore the effects of considering qualitative vs quantitative categorisation of DNA results. We also show the importance of taking into account multiple individual\'s DNA being transferred (such as unknown or wearer DNA), even if the main focus of the evaluation is the activity of one individual.

Failures in mining machinery can abruptly halt mineral production and operations, emphasizing the indispensable role of humans in maintenance and repair operations. Addressing human errors is crucial for ensuring a safe and reliable system, particularly during maintenance activities where accidents frequently occur. This paper focuses on evaluating Human Reliability (HR) to enhance activity implementation effectiveness. Given the challenge of limited and uncertain data on human errors, this study aims to estimate the probability of human errors using Bayesian networks (BN) under uncertain parameters. Applying this approach to assess HR in the maintenance and repair operations of mining trucks at Golgohar Iron Ore Mine in Iran, the study identifies critical factors influencing error occurrence in a fuzzy environment. The results highlight key factors impacting human error and offer insights into estimating HR with minimal human intervention.

Cell therapy, a burgeoning therapeutic strategy, necessitates a scientific regulatory framework but faces challenges in risk-based regulation due to the lack of a global consensus on risk classification. This study applies Bayesian network analysis to compare and evaluate the risk classification strategies for cellular products proposed by the Food and Drug Administration (FDA), Ministry of Health, Labour and Welfare (MHLW), and World Health Organization (WHO), using real-world data to validate the models. The appropriateness of key risk factors is assessed within the three regulatory frameworks, along with their implications for clinical safety. The results indicate several directions for refining risk classification approaches. Additionally, a substudy focuses on a specific type of cell and gene therapy (CGT), chimeric antigen receptor (CAR) T cell therapy. It underscores the importance of considering CAR targets, tumor types, and costimulatory domains when assessing the safety risks of CAR T cell products. Overall, there is currently a lack of a regulatory framework based on real-world data for cellular products and a lack of risk-based classification review methods. This study aims to improve the regulatory system for cellular products, emphasizing risk-based classification. Furthermore, the study advocates for leveraging machine learning in regulatory science to enhance the assessment of cellular product safety, illustrating the role of Bayesian networks in aiding regulatory decision-making for the risk classification of cellular products.

To clarify the complex relationship between the factors causing safety accidents in metallurgical enterprises and predict the risk of accidents in enterprises, a correlation analysis model of the factors causing safety accidents in metallurgical enterprises based on grey Decision-Making Trial and Evaluation Laboratory/Interpretative Structural Modeling (DEMATEL/ISM) was established, and a Bayesian network early warning model was constructed on this basis. The relationship and action path of accident-causing factors in metallurgical enterprises were clarified. The factors were hierarchically divided and a multi-layer hierarchical structure model was established to obtain the neighboring cause, transitional cause, and essential cause of the accident. The results showed that the employee violation rate, the hazardous substances reserves, the toxic gas and dust pollution control compliance rate, the pass rate for equipment maintenance, and the qualification rate of special equipment were the neighboring causes of the accident. The perfection of the safety production management system was the essential cause. The Bayesian network early warning model was applied to the Fuxin Jiuxing Titanium work site. The expected risk probability of an accident was 17.9%, which was in a comparatively safe state (State2). The results obtained by the Bayesian model are consistent with those obtained by AHP and fuzzy comprehensive evaluation method, which proved the accuracy of the early warning model. The Bayesian model can give the risk probability value of the accident and the risk probability value of the accident cause factors at the same time, and include the causal relationship and conditional correlation relationship among the indicator variables in the reasoning process, which can provide targeted technical support for the construction of the emergency system of risk classification management and control.

BACKGROUND: Dimension reduction methods do not always reduce their underlying indicators to a single composite score. Furthermore, such methods are usually based on optimality criteria that require discarding some information. We suggest, under some conditions, to use the joint probability density function (joint pdf or JPD) of p-dimensional random variable (the p indicators), as an index or a composite score. It is proved that this index is more informative than any alternative composite score. In two examples, we compare the JPD index with some alternatives constructed from traditional methods.
METHODS: We develop a probabilistic unsupervised dimension reduction method based on the probability density of multivariate data. We show that the conditional distribution of the variables given JPD is uniform, implying that the JPD is the most informative scalar summary under the most common notions of information. B. We show under some widely plausible conditions, JPD can be used as an index. To use JPD as an index, in addition to having a plausible interpretation, all the random variables should have approximately the same direction(unidirectionality) as the density values (codirectionality). We applied these ideas to two data sets: first, on the 7 Brief Pain Inventory Interference scale (BPI-I) items obtained from 8,889 US Veterans with chronic pain and, second, on a novel measure based on administrative data for 912 US Veterans. To estimate the JPD in both examples, among the available JPD estimation methods, we used its conditional specifications, identified a well-fitted parametric model for each factored conditional (regression) specification, and, by maximizing the corresponding likelihoods, estimated their parameters. Due to the non-uniqueness of conditional specification, the average of all estimated conditional specifications was used as the final estimate. Since a prevalent common use of indices is ranking, we used measures of monotone dependence [e.g., Spearman\'s rank correlation (rho)] to assess the strength of unidirectionality and co-directionality. Finally, we cross-validate the JPD score against variance-covariance-based scores (factor scores in unidimensional models), and the \"person\'s parameter\" estimates of (Generalized) Partial Credit and Graded Response IRT models. We used Pearson Divergence as a measure of information and Shannon\'s entropy to compare uncertainties (informativeness) in these alternative scores.
RESULTS: An unsupervised dimension reduction was developed based on the joint probability density (JPD) of the multi-dimensional data. The JPD, under regularity conditions, may be used as an index. For the well-established Brief Pain Interference Inventory (BPI-I (the short form with 7 Items) and for a new mental health severity index (MoPSI) with 6 indicators, we estimated the JPD scoring. We compared, assuming unidimensionality, factor scores, Person\'s scores of the Partial Credit model, the Generalized Partial Credit model, and the Graded Response model with JPD scoring. As expected, all scores\' rankings in both examples were monotonically dependent with various strengths. Shannon entropy was the smallest for JPD scores. Pearson Divergence of the estimated densities of different indices against uniform distribution was maximum for JPD scoring.
CONCLUSIONS: An unsupervised probabilistic dimension reduction is possible. When appropriate, the joint probability density function can be used as the most informative index. Model specification and estimation and steps to implement the scoring were demonstrated. As expected, when the required assumption in factor analysis and IRT models are satisfied, JPD scoring agrees with these established scores. However, when these assumptions are violated, JPD scores preserve all the information in the indicators with minimal assumption.

UNASSIGNED: Clinical reasoning in veterinary medicine is often based on clinicians\' personal experience in combination with information derived from publications describing cohorts of patients. Studies on the use of scientific methods for patient individual decision making are largely lacking. This applies to the prediction of the individual underlying pathology in seizuring dogs as well. The aim of this study was to apply machine learning to the prediction of the risk of structural epilepsy in dogs with seizures.
UNASSIGNED: Dogs with a history of seizures were retrospectively as well as prospectively included. Data about clinical history, neurological examination, diagnostic tests performed as well as the final diagnosis were collected. For data analysis, the Bayesian Network and Random Forest algorithms were used. A total of 33 features for Random Forest and 17 for Bayesian Network were available for analysis. The following four feature selection methods were applied to select features for further analysis: Permutation Importance, Forward Selection, Random Selection and Expert Opinion. The two algorithms Bayesian Network and Random Forest were trained to predict structural epilepsy using the selected features.
UNASSIGNED: A total of 328 dogs of 119 different breeds were identified retrospectively between January 2017 and June 2021, of which 33.2% were diagnosed with structural epilepsy. An overall of 89,848 models were trained. The Bayesian Network in combination with the Random feature selection performed best. It was able to predict structural epilepsy with an accuracy of 0.969 (sensitivity: 0.857, specificity: 1.000) among all dogs with seizures using the following features: age at first seizure, cluster seizures, seizure in last 24 h, seizure in last 6 month, and seizure in last year.
UNASSIGNED: Machine learning algorithms such as Bayesian Networks and Random Forests identify dogs with structural epilepsy with a high sensitivity and specificity. This information could provide some guidance to clinicians and pet owners in their clinical decision-making process.

The circular economy (CE) is reasoned to organize complex systems supporting sustainable resilience by distinguishing between waste materials and economic growth. This is crucial to the electronic waste (e-waste) industry of developed countries, and e-waste operation management has become their top priority because e-waste contains toxic materials and valuable sources of elements. In the UK, although London Metropolitan city boasts an ambitious sustainable resilience target underlying the context of CE, practical implementation has yet to be feasible, with few investigations detailing if and how the existing target implications enable industrial and social-ecological sectors to continue their performance functionalities in the face of undesired disruptions. In this paper, a dynamic Bayesian Network (dynamic BN) approach is developed to address a range of potential risks. The existing London e-waste operation management is considered as an application of study for sustainable resilience development. Through the utilization of dynamic BN, a comprehensive analysis yields a Resilience Index (RI) of 0.5424, coupled with a StdDev of 0.01350. These metrics offer a profound insight into the intricate workings of a sustainable system and its capacity to swiftly rebound from unexpected shocks and disturbances. This newfound understanding equips policymakers with the knowledge needed to navigate the complexities of sustainable e-waste management effectively. The implications drawn from these in-depth analyses furnish policymakers with invaluable information, enabling them to make judicious decisions that advance the cause of sustainable e-waste management. The findings underscore that the absorptive capacity of a sustainable and resilient e-waste operation management system stands as the foremost defense mechanism against unforeseen challenges. Furthermore, it becomes evident that two pivotal factors, namely \"diversifying the supply chain\" and \"enhancing supply chain transparency,\" play pivotal roles in augmenting the sustainability and resilience of e-waste operation management within the context of London\'s ambitious sustainability targets. These factors are instrumental in steering the trajectory of e-waste management towards a more sustainable and resilient future, aligning with London\'s aspirations for a greener and more eco-conscious future.