Training Database Research Articles

An open codebase for enhancing transparency in deep learning-based breast cancer diagnosis utilizing CBIS-DDSM data

Accessible mammography datasets and innovative machine learning techniques are at the forefront of computer-aided breast cancer diagnosis. However, the opacity surrounding private datasets and the unclear methodology behind the selection of subset images from publicly available databases for model training and testing, coupled with the arbitrary incompleteness or inaccessibility of code, markedly intensifies the obstacles in replicating and validating the model's efficacy. These challenges, in turn, erect barriers for subsequent researchers striving to learn and advance this field. To address these limitations, we provide a pilot codebase covering the entire process from image preprocessing to model development and evaluation pipeline, utilizing the publicly available Curated Breast Imaging Subset of Digital Database for Screening Mammography (CBIS-DDSM) mass subset, including both full images and regions of interests (ROIs). We have identified that increasing the input size could improve the detection accuracy of malignant cases within each set of models. Collectively, our efforts hold promise in accelerating global software development for breast cancer diagnosis by leveraging our codebase and structure, while also integrating other advancements in the field.

Novel IDS Framework: Integrating GWO Feature Selection with KAN Classification

The past few decades have seen significant lifestyle improvements through cutting-edge innovations like AI and fog computing enabled by the World Wide Web, which are crucial for practical applications involving both symmetrical and asymmetrical information distributions. However, these advancements also come with the risk off requent and dangerous cyber-attacks. To mitigate these unwanted malwares, researchers must develop and implement novel intrusion detection systems (IDSs). Challenges such as inadequate accuracy scores due to numerous unnecessary and ineffective features, poor identification of attacks through specific deep learning classification methods, the costly and inefficient use of labeled training databases, and the excessive time required for system development and evaluation hinder there liability and efficiency of IDSs.This paper proposes a hybrid approach that combines the Grey Wolf Optimizer (GWO) for feature selection and the Kolmogorov-Arnold Network(KAN) algorithm for classification to address these challenges.The GWO is utilized to extract important features through meta-heuristic methods, demonstrating its effectiveness in feature selection. The proposed model achieves a high accuracy score, precision, recall, and F1-score of 92.09%, indicating its capability to accurately identify both normal and intrusive activities. The Matthews Correlation Coefficient (MCC) of 0.9019 further validates the model's robustness.

ChatGPT M.D.: Is there any room for generative AI in neurology?

ChatGPT, a general artificial intelligence, has been recognized as a powerful tool in scientific writing and programming but its use as a medical tool is largely overlooked. The general accessibility, rapid response time and comprehensive training database might enable ChatGPT to serve as a diagnostic augmentation tool in certain clinical settings. The diagnostic process in neurology is often challenging and complex. In certain time-sensitive scenarios, rapid evaluation and diagnostic decisions are needed, while in other cases clinicians are faced with rare disorders and atypical disease manifestations. Due to these factors, the diagnostic accuracy in neurology is often suboptimal. Here we evaluated whether ChatGPT can be utilized as a valuable and innovative diagnostic augmentation tool in various neurological settings. We used synthetic data generated by neurological experts to represent descriptive anamneses of patients with known neurology-related diseases, then the probability for an appropriate diagnosis made by ChatGPT was measured. To give clarity to the accuracy of the AI-determined diagnosis, all cases have been cross-validated by other experts and general medical doctors as well. We found that ChatGPT-determined diagnostic accuracy (ranging from 68.5% ± 3.28% to 83.83% ± 2.73%) can reach the accuracy of other experts (81.66% ± 2.02%), furthermore, it surpasses the probability of an appropriate diagnosis if the examiner is a general medical doctor (57.15% ± 2.64%). Our results showcase the efficacy of general artificial intelligence like ChatGPT as a diagnostic augmentation tool in medicine. In the future, AI-based supporting tools might be useful amendments in medical practice and help to improve the diagnostic process in neurology.

Advancing non-alcoholic fatty liver disease prediction: a comprehensive machine learning approach integrating SHAP interpretability and multi-cohort validation.

Non-alcoholic fatty liver disease (NAFLD) represents a major global health challenge, often undiagnosed because of suboptimal screening tools. Advances in machine learning (ML) offer potential improvements in predictive diagnostics, leveraging complex clinical datasets. We utilized a comprehensive dataset from the Dryad database for model development and training and performed external validation using data from the National Health and Nutrition Examination Survey (NHANES) 2017-2020 cycles. Seven distinct ML models were developed and rigorously evaluated. Additionally, we employed the SHapley Additive exPlanations (SHAP) method to enhance the interpretability of the models, allowing for a detailed understanding of how each variable contributes to predictive outcomes. A total of 14,913 participants were eligible for this study. Among the seven constructed models, the light gradient boosting machine achieved the highest performance, with an area under the receiver operating characteristic curve of 0.90 in the internal validation set and 0.81 in the external NHANES validation cohort. In detailed performance metrics, it maintained an accuracy of 87%, a sensitivity of 92.9%, and an F1 score of 0.92. Key predictive variables identified included alanine aminotransferase, gammaglutamyl transpeptidase, triglyceride glucose-waist circumference, metabolic score for insulin resistance, and HbA1c, which are strongly associated with metabolic dysfunctions integral to NAFLD progression. The integration of ML with SHAP interpretability provides a robust predictive tool for NAFLD, enhancing the early identification and potential management of the disease. The model's high accuracy and generalizability across diverse populations highlight its clinical utility, though future enhancements should include longitudinal data and lifestyle factors to refine risk assessments further.

Artificial Neural Network Model for Predicting Carrot Root Yield Loss in Relation to Mechanical Heading

Modelling and predicting agricultural production processes have high cognitive and practical values. Plant growth, the threat of pathogens and pests, and the structure of agricultural machinery treatments are mostly non-linear, measurable processes that can be described mathematically. In this paper, a multiple regression analysis was carried out in the first step to check the non-linearity of the data and yielded a coefficient of determination of R2 = 0.9741 and the coefficient of determination corrected for degrees of freedom was R2adj = 0.9739. An artificial neural network model, called CH-NET, is then presented to predict the yield loss of carrot roots by leaving root mass in the field during harvest at the mechanical heading stage. The proposed network model has an architecture consisting of an input layer, three hidden layers with 12 neurons each, and an output layer with one neuron. Twelve input criteria were defined for the analysis and testing of the network, eight of which related to carrot root parameters and four to the heading machine. The training, testing, and validation database of the CH-NET network consisted of the results of field trials and tests of the operation of the patented (P.242097) root heading machine. The proposed CH-NET neural network model achieved global error (GE) values of 0.0931 t·ha−1 for predicting carrot root yield losses for all twelve criteria adopted. However, when the number of criteria is reduced to eight, the error increased to 0.0991 t·ha−1. That is, the prediction was realised with an accuracy of 90.69%. The developed CH-NET model allows the prediction of economic losses associated with root mass left in the field or contamination of the raw material with undercut leaves. The simulations carried out showed that minimum root losses (0.263 t·ha−1) occur at an average root head projection height of 38 mm and a heading height of 20 mm from the ridge surface.

Creating Future Stations: The Application of Large Generative Models in Metro Design

New technologies, including the use of generative models, are beginning to actively influence design decisions in the decoration of metro stations. The study of the possibilities of using generative artificial intelligence as an art design tool in the decoration of metro stations is relevant, because it can significantly accelerate the design process, reducing time and labour costs for the development and optimization of projects. The novelty of the article lies in the fact that for the first time the process of interaction between the designer and software and hardware of large generative models is considered in detail on the example of designing the interior of a metro station on the theme “ocean”.The problem of creation of original design project in conditions of generative models at the moment is not solvable without direct control on the part of a human, as this technology, for lack of human perception, first of all orientate on expressive means contained in training databases. Therefore, when designing metro stations using generative models, images generated through text often lose the semantic and emotional meanings that the designer put into them. As a result, the generated design project does not fully conform to the designer’s intent and requires further manual adjustment and optimization. Thus, as long as there is a need for conscious and purposeful communication of ideas and feelings through art and design, the problem of endowing a work with meaning using generative models will require human involvement.At the same time, it should be noted that in the artist’s work, the use of generative pre-trained transformers is manifested in the ability to easily create realistic visual effects, to execute images in different styles, focusing on the creative individual manner of famous masters, and to improve design ideas. The advantage of this technology is that it can reduce the designer’s workload while improving the efficiency and accuracy of the design.

Role of inter- and extra-lesion tissue, transfer learning, and fine-tuning in the robust classification of breast lesions

Accurate and unbiased classification of breast lesions is pivotal for early diagnosis and treatment, and a deep learning approach can effectively represent and utilize the digital content of images for more precise medical image analysis. Breast ultrasound imaging is useful for detecting and distinguishing benign masses from malignant masses. Based on the different ways in which benign and malignant tumors affect neighboring tissues, i.e., the pattern of growth and border irregularities, the penetration degree of the adjacent tissue, and tissue-level changes, we investigated the relationship between breast cancer imaging features and the roles of inter- and extra-lesional tissues and their impact on refining the performance of deep learning classification. The novelty of the proposed approach lies in considering the features extracted from the tissue inside the tumor (by performing an erosion operation) and from the lesion and surrounding tissue (by performing a dilation operation) for classification. This study uses these new features and three pre-trained deep neuronal networks to address the challenge of breast lesion classification in ultrasound images. To improve the classification accuracy and interpretability of the model, the proposed model leverages transfer learning to accelerate the training process. Three modern pre-trained CNN architectures (MobileNetV2, VGG16, and EfficientNetB7) are used for transfer learning and fine-tuning for optimization. There are concerns related to the neuronal networks producing erroneous outputs in the presence of noisy images, variations in input data, or adversarial attacks; thus, the proposed system uses the BUS-BRA database (two classes/benign and malignant) for training and testing and the unseen BUSI database (two classes/benign and malignant) for testing. Extensive experiments have recorded accuracy and AUC as performance parameters. The results indicate that the proposed system outperforms the existing breast cancer detection algorithms reported in the literature. AUC values of 1.00 are calculated for VGG16 and EfficientNet-B7 in the dilation cases. The proposed approach will facilitate this challenging and time-consuming classification task.

Robust function guided color encoded single fringe pattern and unwrapping method

Fringe Projection Profilometry (Fringe Projection Profilometry，FPP), renowned for its speed, precision, and automation advantages, has found extensive applications in industries such as mechanical processing and quality inspection. A critical challenge within fringe projection techniques involves ensuring robust unwrapping while minimizing the necessary fringe images. While various unwrapping methods exist to reduce the number of projected images, there is a deficiency in efficient unwrapping methods developed under the guidance of error analysis theory. In this study, we derive a robustness function from FPP's error theory to assess the unwrapping robustness of different encoded images. Using this robustness function, we select highly robust single-color encoded fringe images and achieve unwrapping robustness with minimal training cost based on a single fringe pattern. Furthermore, this unwrapping process does not necessitate the establishment of complex training databases, as the K-nearest neighbor (K-Nearest Neighbor，KNN) algorithm learns the correspondence between image tricolor grayscale values and fringe orders from multi-frequency unwrapping method. Finally, experimental validation through a checkerboard measurement confirms that the proposed method can achieve robust unwrapping with low training costs using only one encoded fringe, thereby affirming the effectiveness of the introduced robustness function and the feasibility of the designed color encoded single fringe pattern unwrapping method.

Dynamic Gradient Boosted Metaheuristic approach for efficient reinforced concrete structure optimization

Structural optimization remains a time-consuming and challenging task, particularly for reinforced concrete structures. Surrogate modeling, a subset of machine learning, has emerged as a promising approach to mitigate the computational burden of structural analysis. However, the use of surrogate modeling also introduces additional challenges, especially the risk of prediction errors due to a small training database. This paper introduces a novel framework, Dynamic Gradient Boosted Metaheuristic Method (DGBM), to address this challenge and enhance optimization efficiency. DGBM integrates Light Gradient Boosted Machine (LightGBM) as a surrogate model for structural analysis, along with a modified Symbiotic Organisms Search (SOS) algorithm, effectively exploring the search space to discover optimal solutions. Its innovative features include LightGBM screening, dynamic training, design modifications, adaptive correction mechanisms, and integration with ETABS. The current study aims to leverage the strengths of machine learning and metaheuristics to significantly reduce the computational cost associated with traditional optimization methods while minimizing the impact of prediction errors. These enhancements refine and expedite the optimization process, making it more applicable to real-world design scenarios. The effectiveness of DGBM is evaluated through case studies on practical reinforced concrete structures, the proposed framework consistently achieves the target cost and delivers results 42.5 % faster in case 1, 57.1 % faster in case 2, and 25 % faster in case 3 compared to the next best method. This advantage translates to faster design iterations and increased efficiency for engineers, empowering them to explore a broader range of design options and achieve optimal solutions for complex structures in a shorter timeframe.

Application of graph neural networks to predict explosion-induced transient flow

We illustrate an application of graph neural networks (GNNs) to predict the pressure, temperature and velocity fields induced by a sudden explosion. The aim of the work is to enable accurate simulation of explosion events in large and geometrically complex domains. Such simulations are currently out of the reach of existing CFD solvers, which represents an opportunity to apply machine learning. The training dataset is obtained from the results of URANS analyses in OpenFOAM. We simulate the transient flow following impulsive events in air in atmospheric conditions. The time history of the fields of pressure, temperature and velocity obtained from a set of such simulations is then recorded to serve as a training database. In the training simulations we model a cubic volume of air enclosed within rigid walls, which also encompass rigid obstacles of random shape, position and orientation. A subset of the cubic volume is initialized to have a higher pressure than the rest of the domain. The ensuing shock initiates the propagation of pressure waves and their reflection and diffraction at the obstacles and walls. A recently proposed GNN framework is extended and adapted to this problem. During the training, the model learns the evolution of thermodynamic quantities in time and space, as well as the effect of the boundary conditions. After training, the model can quickly compute such evolution for unseen geometries and arbitrary initial and boundary conditions, exhibiting good generalization capabilities for domains up to 125 times larger than those used in the training simulations.

Composition driven machine learning for unearthing high-strength lightweight multi-principal element alloys

High-strength, lightweight alloys are highly desired in the aerospace, automotive and marine industries. The optimization of such alloys is a multifaceted process, characterized by the consideration of diverse objectives and constraints. Various optimization methodologies exist, spanning from intricate first-principle approaches to the application of sophisticated machine learning algorithms. These algorithms might incorporate input features encompassing elemental composition, microstructural attributes, and thermodynamic properties to enhance prediction accuracies. In this work, we aim to streamline this complexity by employing solely the alloy's elemental composition as the input feature for the machine learning algorithm, improving the hardness while reducing the density of the alloy. We have curated a comprehensive database comprising 544 multi-principal element alloys and developed a robust surrogate model based on these compositions. This composition-driven model is subsequently coupled with principal component analysis (PCA) to facilitate the selection process. Remarkably, through a mere three iterations involving 14 samples, we successfully identified an alloy with an effective specific hardness surpassing the training database maximum by 8.6 %. The proposed composition-driven machine learning delineates a simplified approach for conducting optimization across multiple target material properties.

Transferable and data efficient metamodeling of storm water system nodal depths using auto-regressive graph neural networks

Storm water systems (SWSs) are essential infrastructure providing multiple services including environmental protection and flood prevention. Typically, utility companies rely on computer simulators to properly design, operate, and manage SWSs. However, multiple applications in SWSs are highly time-consuming. Researchers have resorted to cheaper-to-run models, i.e. metamodels, as alternatives of computationally expensive models. With the recent surge in artificial intelligence applications, machine learning has become a key approach for metamodelling urban water networks. Specifically, deep learning methods, such as feed-forward neural networks, have gained importance in this context. However, these methods require generating a sufficiently large database of examples and training their internal parameters. Both processes defeat the purpose of using a metamodel, i.e., saving time. To overcome this issue, this research focuses on the application of inductive biases and transfer learning for creating SWS metamodels which require less data and retain high performance when used elsewhere. In particular, this study proposes an auto-regressive graph neural network metamodel of the Storm Water Management Model (SWMM) from the Environmental Protection Agency (EPA) for estimating hydraulic heads. The results indicate that the proposed metamodel requires a smaller number of examples to reach high accuracy and speed-up, in comparison to fully connected neural networks. Furthermore, the metamodel shows transferability as it can be used to predict hydraulic heads with high accuracy on unseen parts of the network. This work presents a novel approach that benefits both urban drainage practitioners and water network modeling researchers. The proposed metamodel can help practitioners on the planning, operation, and maintenance of their systems by offering an efficient metamodel of SWMM for computationally intensive tasks like optimization and Monte Carlo analyses. Researchers can leverage the current metamodel’s structure for developing new surrogate model architectures tailored to their specific needs or start paving the way for more general foundation metamodels of urban drainage systems.

Forme Fruste Keratoconus Detection with OCT Corneal Topography Using Artificial Intelligence Algorithms.

Distinguishing early Keratoconus (KC) from normal corneas is challenging owing to their striking similarities. The aim of our study was to identify discriminating parameters to differentiate a normal cornea from a Form Fruste Keratoconus (FFKC) with the Swept-Source (SS) OCT-topography CASIA 2 (Tomey,Japan) using machine learning artificial intelligence algorithms. The study was monocentric, carried out in Bordeaux. This was a retrospective study, case control. Three groups were included: KC group (108 eyes), FFKC (88 eyes) and normal corneas (162 eyes).The data were analyzed and processed using the Dataiku data science platform. Machine learning models (Random Forest, Logistic Regression) were used to develop a multiclass classifier for automated early KC detection. The models were trained using a training database and tested using a test database. Then algorithms were compared to the Ectasia Screening Index (ESI), which is an OCT-topography inherent screening score for ectasia. The Logistic Regression (LR), and Random Forest (RF) detected FFKC with an AUC of 0,99, and 0,98 respectively. The sensitivities of LR (100%), RF (84%) were better than the ESI (28%) for the diagnosis of FFKC. However, ESI has a maximum specificity (100%) compared to the RL (100%) and 90% for RF. This study identified discriminating topographic parameters to be considered in refractive surgery screening on SS-OCT CASIA 2. We developed an algorithm capable of classifying normal eyes versus FFKC cases, with improved performance compared to the ESI score.

A Kronecker product accelerated efficient sparse Gaussian Process (E-SGP) for flow emulation

In this paper, we introduce an efficient sparse Gaussian process (E-SGP) for the surrogate modelling of fluid mechanics. This novel Bayesian machine learning algorithm enables efficient model training using databases of different structures. It extends the approximated sparse GP algorithm by combining the concept of efficient GP (E-GP) and variational energy free sparse Gaussian processes (VEF-SGP). The developed E-SGP approach leverages the arbitrariness of inducing points and the monotonically increasing nature of the objective function with respect to the number of inducing points in VEF-SGP. By specifying the inducing points on the orthogonal grid/input subspace and using the Kronecker product, E-SGP significantly enhances computational efficiency without imposing constraints on the covariance matrix or increasing the number of parameters that need to be optimised during training.The E-SGP algorithm outperforms E-GP not only in terms of scalability but also in model quality, as evidenced by mean standardised logarithmic loss (MSLL). The computational complexity of E-GP grows cubically with an increasing structured training database, while E-SGP maintains computational efficiency as the model resolution (i.e., the number of inducing points) remains the same. The examples show that E-SGP produces more accurate predictions compared to E-GP when model resolutions are similar. In the application to a partially structured database, E-GP benefits from more training data but comes with higher computational demands, while E-SGP achieves a comparable level of accuracy but is more computationally efficient, making E-SGP a potentially preferable choice for fluid mechanics. Furthermore, E-SGP can produce more reasonable estimates of model uncertainty, whilst E-GP is more likely to produce over-confident predictions. In the application to partially structured databases, the performance of E-SGP is also compared to the structured Gaussian processes latent variable model (SGPLVM). In this case, E-SGP and SGPLVM both exhibit robust performance. However, E-SGP demonstrates a slight advantage over SGPLVM, particularly in terms of MSLL, indicating that E-SGP is better at producing trustworthy and accurate estimates of the uncertainty in the predictions. . While SGPLVM also improves computational efficiency by exploiting the Kronecker product and a small number of inducing points, E-SGP's orthogonal grid specification of inducing points further enhances its scalability and robustness, making it a superior choice for complex fluid mechanics problems with partially structured data.

Machine-Learning-Accelerated DFT Conformal Sampling of Catalytic Processes.

Computational modeling of catalytic processes at gas/solid interfaces plays an increasingly important role in chemistry, enabling accelerated materials and process optimization and rational design. However, efficiency, accuracy, thoroughness, and throughput must be enhanced to maximize its practical impact. By combining interpolation of DFT energetics via highly accurate Machine-Learning Potentials with conformal techniques for building the training database, we present here an original approach (that we name Conformal Sampling of Catalytic Processes, CSCP), to accelerate and achieve an accurate and thorough sampling of novel systems by exporting existing information on a worked-out case. We use methanol decomposition (of interest in the field of hydrogen production and storage) as a test catalytic reaction. Starting from worked-out Pt-based systems, we show that after only two iterations of active-learning CSCP is able to provide reaction energy diagrams for a set of 7 diverse systems (Pd, Ni, Au, Ag, Cu, Co, Fe) leading to DFT-accuracy-level predictions. Cases exhibiting a change in adsorption sites and mechanisms are also successfully reproduced as tests of catalytic path modification. The CSCP approach thus offers itself as an operative tool to fully take advantage of accumulated information to achieve high-throughput sampling of catalytic processes.

Moho depth model of South America from a machine learning approach

Crustal structure models play an important role in the characterization of seismogenic zones, in the regional delimitation of geological provinces and particularly, in understanding the genesis and evolution of sedimentary basins. Despite the increasing number of new seismic surveys and seismographic networks in South America, crustal thickness measurements are still scarce and irregularly sampled, reducing the resolution of crustal model maps. To overcome these challenges, a novel approach based on machine learning techniques is proposed, in order to explore higher resolution gravity datasets in the interpolation of crustal thickness measurement points obtained from previous seismic/seismological compilations. The algorithm used in this study is based on Gaussian processes prediction methods, which allowed inferring the depth of Moho to South America. The prediction error of the model obtained from the training and testing database was 3.48 km, which is compatible with the uncertainties derived from the H-k stacking analysis. The depth range varied from 69.8 km beneath the Andes to 4.3 km in oceanic regions. The average Moho depth for the South American Platform is 39.1 km, allowing a spatial correlation of deeper and shallower Moho regions with different types of continental basins. Compared to other models, the model resulting from this study presents fine-scale features highlighting the limits of the main tectonic domains and a good agreement with the suture zones. Overall, this study demonstrates the potential of applying machine learning tools in crustal-scale imaging using sparse datasets, providing new advances in Moho modeling of the South America, as well as new perspectives on its the history and tectonic evolution.

Research on Intelligent Ventilation System of Metal Mine Based on Real-Time Sensing Airflow Parameters with a Global Scheme

In ventilation systems of metal mines, the real-time measurement of the airflow field and a reduction in pollutants are necessary for clean environmental management and human health. However, the limited quantitative data and expensive detection technology hinder the accurate assessment of mine ventilation effectiveness and safety status. Therefore, we propose a new method for constructing a mine intelligent ventilation system with a global scheme, which can realize the intelligent prediction of unknown points in the mine ventilation system by measuring the airflow parameters of multiple known points. Firstly, the nodal wind pressure method combined with the Hardy–Cross iterative algorithm is used to solve the mine ventilation network, and the airflow parameters under normal operation and extreme working conditions are simulated, based on which an intelligent ventilation training database is established. Secondly, we compared the airflow parameter prediction ability of three different machine learning models with different neural network models based on the collected small-sample airflow field dataset of a mine roadway. Finally, the depth learning method is optimized to build the intelligent algorithm model of the mine ventilation system, and a large number of three-dimensional simulation data and field measurement data of the mine ventilation system are used to train the model repeatedly to realize the intelligent perception of air flow parameters of a metal mine ventilation network and the construction of an intelligent ventilation system. The results show that the maximum error of a single airflow measurement point is 1.24%, the maximum overall error is 3.25%, and the overall average error is 0.51%. The intelligent algorithm has a good model training effect and high precision and can meet the requirements of the research and application of this project. Through case analysis, this method can predict the airflow parameters of any position underground and realize the real-time control of mine safety.

Multivitamin Compliance Reduces Injuries of Female Recruits at Air Force Basic Training: A Randomized Controlled Cohort Study.

Department of the Air Force basic military training (BMT) is the 7.5-week initial entry training for all enlisting U.S. Air Force and Space Force recruits. Overuse musculoskeletal injuries (MSKI) during training threaten success in BMT, and trainees with nutritional deficiencies are at higher risk. Several efforts are made at BMT to mitigate these issues to help trainees graduate on time, such as the distribution of multivitamins (MVI) to female recruits. However, trainee compliance with the prescribed MVI has been reported to be low, calling into question the effectiveness of this intervention. This study evaluated the effect of briefing modality offered to these trainees at the time of distribution on educational effectiveness, MVI compliance, and injury rates. This was a randomized controlled cohort study of BMT female trainees. Subjects included all female trainees of a single iteration of BMT concurrently completing training within the same squadron. At the outset of training, cohorts were randomly educated in one of two ways regarding the provided MVI. The control cohort (n = 80) was presented with the standard educational video created by BMT-registered dietitians, while the test cohort (n = 79) was presented the video and then also briefed in-person by a group of subject matter experts. At the end of BMT, the trainees completed a questionnaire regarding their knowledge, confidence, and compliance with the MVI. Additionally, a routinely collected training database was queried and analyzed for differences in injuries and training outcomes using the chi-squared test with significance set to α ≤ 0.05. Although both cohorts found their educational presentation useful, the "in-person" cohort reported significantly higher compliance rates and were found to be diagnosed with MSKI and bone stress injuries at significantly lower rates (43% and 66% relative risk reductions, respectively). Education to female BMT trainees regarding the MVI, which included a face-to-face discussion with subject matter experts, resulted in greater compliance and reduced incidence of MSKI and bone stress injury, compared to a video prepared by registered dietitians alone. Such educational approaches should continue to be strived for in this and other military population health measures. This study's implication of MVI efficacy for injury risk reduction in female trainees should be further studied, verified, and improved upon in this and other populations.

Prediction and analysis of damage to RC columns under close-in blast loads based on machine learning and Monte Carlo method

Rapid prediction and quantitative assessment of the damage of reinforced concrete (RC) columns under blast loads are challenging and crucial issues. The key parameters affecting the anti-blast capacity of RC columns are coupled with failure modes. In this study, machine learning (ML) and Monte Carlo (MC) simulations are employed to investigate the damage of RC columns subjected to blast loads. 257 data collected from existing experimental and numerical studies are utilized to establish a database for model training and testing. The damage indexes of columns are predicted using eight ML models with eight input features. The predictive capacity of each model is characterized by eight evaluation indexes through MC simulations. The CatBoost model is identified as the optimal model based on the Analytic Hierarchy Process (AHP). Additionally, the CatBoost model is explained using the SHapley Additive exPlanations (SHAP) method, and the influence of axial compression ratio on column damage is determined to be intricate. The coupling relationship between the axial compression ratio and the scale distance of the column is analyzed. Finally, a zonal diagram is developed. This diagram can be utilized to assess the damage of the RC column quickly and efficiently.

Atomic cluster expansion potential for large scale simulations of hydrocarbons under shock compression.

We present an Atomic Cluster Expansion (ACE) machine learned potential developed for high-fidelity atomistic simulations of hydrocarbons, targeting pressures and temperatures near and above supercritical fluid regimes for molecular fluids. A diverse set of stoichiometries were covered in training, including 1:0 (pure carbon), 1:4 (methane), and 1:1 (benzene), and rich bonding environments sampled at supercritical temperatures, hydrogen rich, reactive mixtures where metastable stoichiometries arise, including 1:2 (ethylene) and 1:3 (ethane). A high-fidelity training database was constructed by performing large-scale quantum molecular dynamic simulations [density functional theory (DFT) MD] of diamond, graphite, methane, and benzene. A novel approach to selecting structures from DFT MD is also presented, which allows for the rapid selection of unique DFT MD frames from complex trajectories. Comparisons to DFT and experimental data demonstrate that the presented ACE potential accurately reproduces isotherms, carbon melting curves, radial distribution functions, and shock Hugoniots for carbon and hydrocarbon systems for pressures up to 100GPa and temperatures up to 6000K for hydrocarbon systems and up to 9000K for pure carbon systems. This work delivers a potential that can be used for accurate, large-scale simulations of shocked hydrocarbons and demonstrates a methodology for fitting and validating machine learning interatomic potentials to complex molecular environments, which can be applied to energetic materials in future works.