Sepsis poses a serious threat to patient survival, making timely risk assessment crucial. Predicting in-hospital mortality based on clinical indicators can aid in making better clinical decisions. Previous studies have focused on classifier selection but lacked a comprehensive analysis of feature selection and data preprocessing. This study optimized machine learning models for sepsis mortality prediction by: (1) comprehensively comparing feature selection and classification methods to identify the best combination, (2) building a high-performing model with fewer features, and (3) identifying key clinically relevant indicators.Methods: Using the MIMIC-III sepsis cohort, we conducted a comprehensive analysis to determine the optimal model, including data preprocessing, data balance, classifier selection, and feature selection. Feature importance was further analyzed to identify the key predictors of in-hospital mortality.Results: The proposed Synthetic Minority Oversampling Technique-Random Forest Recursive Feature Elimination-Extreme Gradient Boosting (SMOTE-(RF-RFE)-XGB) model achieved high predictive performance with a mean Area Under the Curve (AUC) of 0.8507, while reducing the number of features from 78 to 39. Compared to other feature selection methods evaluated in this study and those reported in related literature, Random Forest Recursive Feature Elimination (RF-RFE) offers the best trade-off between accuracy, feature compactness, and stability. Additionally, feature importance rankings consistently identified Acute Physiology Score III (APS III), Ventilation on First Day, and Depression as the top three most influential predictors, besides the Length of Stay in ICU and Hospital.Conclusions: This study addresses key gaps by conducting a comprehensive evaluation of classifiers and feature selection methods for predicting in-hospital mortality in patients with sepsis. The proposed SMOTE-(RF-RFE)-XGB model achieved a high predictive performance and stability with a compact feature set. APS III, Ventilation on First Day, and Depression were consistently identified as key predictors besides Length of Stay in ICU and Hospital.
Pradyut Anand, Surya Dev Singh, Suresh Pratap
et al.
Abstract The current study outlines an integrated experimental and data-driven methodology of the development of sustainable high-performance concrete, using a hybrid low-carbon binder, which includes fly ash (FA), ground granulated blast-furnace slag (GGBS), thermally treated coir biomass (TTCB) and graphene nanoplatelets (GNPs). M40 grade concrete mixtures were prepared by systematic variation of supplementary cementitious materials (SCMs, 30% of the total binder), TTCB (5% to 10%) and GNPs dosage (0.08% to 0.12%). Mechanical properties were evaluated at 7 and 28 days, rapid chloride permeability (RCPT), water uptake and residual strength after exposure to 300 °C. The optimised mix delivered a compressive strength of 55 MPa at 28 days, which is approximately 23% greater than the control mix (44–45 MPa at 28 days), and chloride permeability 505 C (42% lower) and water absorption 2.8% (40% lower), respectively. Retention of strength following exposure to 300 °C was above 80%, which means that the thermal stability was improved. Microstructural examinations confirmed refined pore structure, lower content of portlandite and enhanced interfacial bonding. The 60 experimental observations based on replicated specimens in 10 mix designs were trained on Random Forest, XGBoost and CNN LSTM models. XGBoost had the best predictive accuracy (R2 > 0.95 to predict strength), and the permutation-importance analysis revealed TTCB content and SCMs balance to be the most important predictors. Multi-objective optimisation (NSGA-II and MOEA/D) was used to produce trade-offs between strength, durability, embodied CO2 and cost within the constrained experimental space. The suggested surrogate-assisted optimisation model offers a repeatable approach to eco-efficient concrete mix design at limited laboratory conditions.
Wheeled bipedal robots are promising for industrial mobility because they combine tight turning, agile balancing, and efficient rolling. Their inherently unstable and underactuated dynamics make reliable reference tracking challenging, particularly in the presence of sustained external disturbances and modeling errors. This paper presents a systematic modeling and control study using a three-degrees-of-freedom sagittal plane representation derived from the original six-degrees-of-freedom dynamics. Two linear tracking controllers are designed and compared: a full state feedback tracking controller and a linear quadratic servo controller with integral action. Practical performance is validated through real-time hardware in the loop simulation, where the controller runs on embedded hardware and the plant is executed on a real-time target including discrete time-sampling effects and analog input output communication noise associated with signal transmission. The results show that both controllers achieve stabilization, while the comparative HILS results reveal a trade-off rather than a uniformly superior controller. The full state feedback controller often yields lower finite-horizon position tracking errors, whereas the linear quadratic servo controller provides tighter body-pitch regulation and the more reliable removal of steady-state offset under sustained constant disturbances. These results demonstrate the feasibility of optimal servo control on cost-effective embedded platforms and indicate that controller selection should depend on the desired balance, considering tracking accuracy, disturbance rejection, convergence behavior, and actuator usage.
Materials of engineering and construction. Mechanics of materials, Production of electric energy or power. Powerplants. Central stations
Ariadna Fosch, Alberto Aleta, Roger Cremades
et al.
Recent global food trade disruptions have evidenced how local shocks can cascade into global security threats. While the capacity of food systems to absorb spillovers depends heavily on its underlying trade networks, few studies quantify how their temporal evolution reshapes systemic vulnerability over time. Here, we evaluate how changes in global connectivity from 1986 to 2022 reshaped responses to production shocks. Using FAO data, we built yearly multiplex representations of the food trade system and quantified robustness through a stochastic shock-propagation model with dynamic export bans. We find that while increasing globalization intensified inter-dependencies and amplified cascades, robustness trends remain heterogeneous. Grain trade has become more decentralized and resilient to targeted shocks; conversely, Animal and Vegetable Fats exhibit growing centralization and fragility around key exporters like Indonesia and Malaysia. These structural transformations caused diverging shifts in systemic vulnerability, disproportionately threatening already vulnerable regions such as Africa and Southern Asia.
Jing Luo,1 Dan Zhang,1 Miaomiao Zhang,1 Yiqiang Chen,1 Yi Ding2 1Department of Chinese Medicine, Shenzhen Qianhai Shekou Free Trade Zone Hospital, Shenzhen, Guangzhou, People’s Republic of China; 2Department of Physical Medicine and Rehabilitation, Changsha Social Work College, Changsha, Hunan, People’s Republic of ChinaCorrespondence: Yi Ding, Department of Physical medicine and rehabilitation, Changsha Social Work College, Changsha, Hunan, People’s Republic of China, 410004, Email zyk20240403@163.comBackground: Post-infectious cough (PIC) is a leading cause of chronic cough in children, often persisting after respiratory infections and significantly impairing quality of life. Current therapies, such as montelukast sodium (MAS), offer only partial symptom relief and do not target the underlying inflammatory and microbiota-driven mechanisms. Emerging evidence suggests that the lung–gut axis, ERK pathway activation, and cytokine–microbiota interactions are central to PIC pathogenesis. Morusin, a prenylated flavonoid from Morus alba, possesses anti-inflammatory and barrier-protective activities and may uniquely modulate both ERK signaling and gut microbiota, offering mechanistic advantages over conventional treatments.Methods: A juvenile rat model of PIC was induced by smoke exposure, lipopolysaccharide nasal instillation, and capsaicin atomization. Rats were assigned to control, model, morusin, or MAS groups. Physiological outcomes, histology, and immunostaining were assessed, including body weight, airway resistance, goblet cells, cytokines (IL-4, IL-6, IL-10), and phosphorylated ERK1/2 (p-ERK1/2) in lung and colon tissues. Gut microbiota was profiled via 16S rRNA sequencing, with correlation analyses linking microbial changes to cytokine and signaling profiles.Results: Morusin improved systemic parameters (body weight, salivary flow, skin hydration), reduced airway hyperreactivity, and normalized anxiety-like behaviors, effects not observed with MAS. Both morusin and MAS reduced lung goblet cell hyperplasia and inflammatory cytokines, but only morusin suppressed p-ERK1/2 in both lung and colon tissues and reshaped the gut microbiota. Morusin enriched beneficial genera (Lactobacillus, Akkermansia) and reduced pro-inflammatory taxa (Ruminococcus, Lachnospiraceae_NK4A136_group). Correlation analyses confirmed strong links between microbial shifts, cytokine balance, and ERK modulation.Conclusion: Morusin alleviates PIC through systemic, mucosal, and behavioral improvements, combined with unique modulation of gut microbiota and ERK signaling across the lung–gut axis. These findings highlight morusin’s novel mechanistic advantage over MAS and support its potential as a translational therapy for pediatric PIC.Keywords: morusin, post-infectious cough, air way inflammation, p-ERK, gut microbiota
Abdul Awol Rabby, Imon Hasan Bhuiyan, Uzzal Al Aziz
et al.
The depletion of natural aggregates and the rising volume of construction and demolition waste have made sustainable alternatives in concrete production essential. Recycled coarse aggregate (RCA) helps reduce landfill use and reliance on natural resources, but its variable quality raises concerns about structural reliability. Therefore, evaluating the balance between mechanical performance and economic feasibility is crucial for promoting sustainable construction. In this study, concrete mixes were designed with varying RCA replacement levels (0%, 50%, 70%, and 100%) compared to natural coarse aggregate (NCA). A controlled mix proportion of 1:1.5:3 (cement:sand:aggregate) with a water–cement ratio of 0.5 and 1% superplasticizer was employed. To ensure consistency, aggregates were classified into 19 mm (40%), 12.5 mm (30%), and 9.5 mm (30%) gradations. Cylindrical specimens (100 × 200 mm) were cast and cured, followed by compressive strength testing at 7, 14, and 28 days in accordance with ASTM C39. The total material cost for each mix was also computed on a per-cubic-meter basis to assess economic implications alongside strength performance. The results highlight a clear strength–cost trade-off. The control mix (100% NCA) achieved 26.5 MPa at 28 days, while the 50% RCA mix showed only a 3.92% reduction (25.5 MPa) with 18.4% lower cost. At 70% RCA, compressive strength dropped by 10.42% (24 MPa) with a 28.29% cost reduction, whereas 100% RCA replacement resulted in a severe 59.63% strength loss (16.5 MPa) despite maximum cost savings (49.07%). These findings establish 50% RCA + 50% NCA as the most rational compromise, offering structural adequacy with substantial cost efficiency, while also reinforcing RCA’s role in sustainable construction.
Federated Learning (FL), a key paradigm in privacy-preserving and distributed machine learning (ML), enables collaborative model training across decentralized data sources without requiring raw data exchange. FL enables collaborative model training across decentralized data sources while preserving privacy. However, selecting appropriate clients remains a major challenge, especially in heterogeneous environments with diverse battery levels, privacy needs, and learning capacities. In this work, a centralized reward-based payoff strategy (RBPS) with cooperative intent is proposed for client selection. In RBPS, each client evaluates participation based on locally measured battery level, privacy requirement, and the model’s accuracy in the current round computing a payoff from these factors and electing to participate if the payoff exceeds a predefined threshold. Participating clients then receive the updated global model. By jointly optimizing model accuracy, privacy preservation, and battery-level constraints, RBPS realizes a multi-objective selection mechanism. Under realistic simulations of client heterogeneity, RBPS yields more robust and efficient training compared to existing methods, confirming its suitability for deployment in resource-constrained FL settings. Experimental analysis demonstrates that RBPS offers significant advantages over state-of-the-art (SOA) client selection methods, particularly those relying on a single selection criterion such as accuracy, battery, or privacy alone. These one-dimensional approaches often lead to trade-offs where improvements in one aspect come at the cost of another. In contrast, RBPS leverages client heterogeneity not as a limitation, but as a strategic asset to maintain and balance all critical characteristics simultaneously. Rather than optimizing performance for a single device type or constraint, RBPS benefits from the diversity of heterogeneous clients, enabling improved accuracy, energy preservation, and privacy protection all at once. This is achieved by dynamically adapting the selection strategy to the strengths of different client profiles. Unlike homogeneous environments, where only one capability tends to dominate, RBPS ensures that no key property is sacrificed. RBPS thus aligns more closely with real-world FL deployments, where mixed-device participation is common and balanced optimization is essential.
This paper presents an experimental and numerical investigation of a 254 mm resin-printed propeller operating at rotational speeds between 3000 and 9000 RPM. Propeller thrust and torque were measured using a six-degree-of-freedom load cell, while acoustic data were captured with a microphone positioned three times the propeller diameter from the center. To complement the experimental analysis, computational simulations were conducted using ANSYS Fluent with the detached eddy simulation (DES) model, the Ffowcs-Williams and Hawkings (FW-H) model, and a transient flow solver. The figure of merit (FM) results show that the resin propeller slightly outperforms two commercial counterparts with a marginal difference between the wood and resin propellers. Additionally, the resin propeller demonstrates better noise performance, exhibiting the lowest primary tonal noise, broadband noise, and overall sound pressure level (OASPL), with minimal differences between the two commercial counterparts. ANSYS Fluent simulations predict thrust and torque within a 10% error margin, showing particularly accurate results for primary tonal noise. A new trade-off index is proposed to assess the balance between propeller performance and aeroacoustics, revealing distinct trends compared to traditional metrics. Furthermore, aerodynamic phenomena such as flow separation on the leading edge near the tip, flow separation behind the middle trailing edge, and vortex interactions at the root are identified as key contributors to tonal and broadband noise. These findings provide valuable insights into propeller design and aeroacoustic optimization.
Caison Ramos, Gustavo Marchesan, Ghendy Cardoso
et al.
Isolated microgrids with green hydrogen storage offer a promising solution for supplying electricity to remote communities where conventional grid expansion is infeasible. Designing such systems requires balancing two conflicting objectives: minimizing installation and operation costs while maximizing supply reliability. This paper proposes a multi-objective optimization methodology, based on the Non-dominated Sorting Genetic Algorithm II, to determine the optimal sizing of multiple microgrid components. This sizing explicitly addresses both the power capacities (kW) (for photovoltaic panels, wind turbines, electrolyzers, and fuel cells) and the energy storage capacities (kWh and kg) (for batteries and hydrogen tanks, respectively), aiming to generate Pareto-optimal solutions that explore this trade-off. The proposed method evaluates the trade-off by minimizing two objectives: the Net Present Value, which includes investment, replacement, and maintenance costs, and the total expected interruption hours, derived from an hourly energy balance analysis. The methodology’s effectiveness is validated using four distinct case studies. Three of these are based on real locations with specific load profiles and climate data. To test the method’s robustness, a fourth case study uses a fictitious load profile, designed with pronounced seasonal variations and a clear distinction between weekday and weekend consumption. Our results demonstrate the method’s ability to identify efficient hybrid renewable topologies combining photovoltaic and/or wind generation, batteries, and hydrogen systems (electrolyzer, storage tank, and fuel cell). The obtained cost–reliability curves provide practical decision-support tools for system planners.
New methods are needed to monitor environmental treaties, like the Montreal Protocol, by reviewing large, complex customs datasets. This paper introduces a framework using unsupervised machine learning to systematically detect suspicious trade patterns and highlight activities for review. Our methodology, applied to 100,000 trade records, combines several ML techniques. Unsupervised Clustering (K-Means) discovers natural trade archetypes based on shipment value and weight. Anomaly Detection (Isolation Forest and IQR) identifies rare "mega-trades" and shipments with commercially unusual price-per-kilogram values. This is supplemented by Heuristic Flagging to find tactics like vague shipment descriptions. These layers are combined into a priority score, which successfully identified 1,351 price outliers and 1,288 high-priority shipments for customs review. A key finding is that high-priority commodities show a different and more valuable value-to-weight ratio than general goods. This was validated using Explainable AI (SHAP), which confirmed vague descriptions and high value as the most significant risk predictors. The model's sensitivity was validated by its detection of a massive spike in "mega-trades" in early 2021, correlating directly with the real-world regulatory impact of the US AIM Act. This work presents a repeatable unsupervised learning pipeline to turn raw trade data into prioritized, usable intelligence for regulatory groups.
This paper studies learning in markets with aggregate uncertainty about whether trade is efficient. A long-lived seller offers prices to buyers, who are short-lived and arrive according to a Poisson process. A hidden state determines whether the buyers' common value exceeds the seller's reservation value. All parties observe noisy, private signals about the state. With small intertemporal frictions and when the seller has commitment power, the seller waits for a buyer with the most favorable signal to arrive up to an exit time that depends on the seller's private information. This strategy profile maximizes both the seller's profit and the expected surplus. Without commitment, the commitment profit is unattainable. Instead, there is an equilibrium in which the seller also waits for a buyer with the most favorable signal, but, relative to the commitment case, the seller exits inefficiently late, and the trade probability is inefficiently high.
Industry is the basis of economic development, therefore, studying the real sector allow us to assess the physical scale of the economy of individual countries and the world as a whole. The industrial production structural distribution by countries and macro-regions reflects the real balance of power in the global arena. Retrospective analysis allows us to see this process in dynamics, which contributes to the correct construction of models of further international cooperation, foreign trade and infrastructural development in this direction. The purpose of the study is to assess the dynamics of the industrial production shift from West to East in recent decades. The tabular, graphical and descriptive methods were used, and it allowed us to visualize the indicators retrospective analysis results. In the course of the work it has been proved that by now there is parity in the level of industrial production between the countries of the collective West and East. It makes it possible to change the term “developed countries” meaning and extend it not only to European and North American countries, but also to a number of Asian and Latin American countries due to the ever-growing importance of these countries’ industry on a global scale.
This research paper discusses the impacts on the global economy and international market affected by the US and China trade war, which started in 2018-19. Led by the previous president of the US, Donald Trump, succeeded by the current president Joe Biden, trade protectionism was brought out to restrict Chinese exports to the US. Due to political, social, economic, and many other factors, both US and China ended up imposing additional tariffs on each other’s imports and setting up more and more restrictions on the international market. These imposed trade barriers between the two major economies in the world significantly influenced the two countries themselves, other bystander economies, and the international market balance. The paper discusses and reveals how the economic conflict affects US and China, both consequences and benefits, and how some countries found opportunities from this conflict, and some resulted in losses. Suggestions for possible solutions which each government can take are also being explained.
This paper documents a new feature of international trade: the increase in the share of low-value transactions in the total volume of transactions. Using Spanish data, we show that the share of low-value transactions in the total number of transactions increased from 9% to 61% in exports and from 14% to 54% in imports between 1997 and 2023. The increase in the number of low-value trade transactions is explained by the rise of e-commerce and direct-to-customer sales facilitated by online retail platforms, and the fast-fashion strategy followed by clothing firms.
Kallampally and Tewari showed in 2016 that there can be a trade-off between determinism and time in space-bounded computations. This they did by describing an unambiguous non-deterministic algorithm to solve Directed Graph Reachability that requires O(log^2 n) space and simultaneously runs in polynomial time. Savitch's 1970 algorithm that solves the same problem deterministically also requires O(log^2 n) space but doesn't guarantee polynomial running time and hence the trade off. We describe a new problem for which we can show a similar trade off between determinism and time. We consider a collection P of f directed paths. We show that the problem of finding reachability from one vertex to another in the union G of these path graphs via a path that switches amongst the paths in P at most k times can be solved in O(klog f+log n) space but the algorithm doesn't guarantee polynomial runtime. On the other hand, we also show that the same problem can be solved by an unambiguous non-deterministic algorithm that simultaneously runs in O(klog f+log n) space and polynomial time. Since these two algorithms are not dependent on Savitch, therefore this example sheds new light on how such a trade off between determinism and time happens in space-bounded computations and makes the phenomenon less elusive.
Thiago C. Silva, Paulo V. B. Wilhelm, Diego R. Amancio
This study examines the effects of de-globalization trends on international trade networks and their role in improving forecasts for economic growth. Using section-level trade data from nearly 200 countries from 2010 to 2022, we identify significant shifts in the network topology driven by rising trade policy uncertainty. Our analysis highlights key global players through centrality rankings, with the United States, China, and Germany maintaining consistent dominance. Using a horse race of supervised regressors, we find that network topology descriptors evaluated from section-specific trade networks substantially enhance the quality of a country's GDP growth forecast. We also find that non-linear models, such as Random Forest, XGBoost, and LightGBM, outperform traditional linear models used in the economics literature. Using SHAP values to interpret these non-linear model's predictions, we find that about half of most important features originate from the network descriptors, underscoring their vital role in refining forecasts. Moreover, this study emphasizes the significance of recent economic performance, population growth, and the primary sector's influence in shaping economic growth predictions, offering novel insights into the intricacies of economic growth forecasting.
This paper explores the novel deep learning Transformers architectures for high-frequency Bitcoin-USDT log-return forecasting and compares them to the traditional Long Short-Term Memory models. A hybrid Transformer model, called \textbf{HFformer}, is then introduced for time series forecasting which incorporates a Transformer encoder, linear decoder, spiking activations, and quantile loss function, and does not use position encoding. Furthermore, possible high-frequency trading strategies for use with the HFformer model are discussed, including trade sizing, trading signal aggregation, and minimal trading threshold. Ultimately, the performance of the HFformer and Long Short-Term Memory models are assessed and results indicate that the HFformer achieves a higher cumulative PnL than the LSTM when trading with multiple signals during backtesting.
Julia Adamska, Łukasz Bielak, Joanna Janczura
et al.
Multivariate modelling of economics data is crucial for risk and profit analyses in companies. However, for the final conclusions, a whole set of variables is usually transformed into a single variable describing a total profit/balance of company’s cash flows. One of the possible transformations is based on the product of market variables. Thus, in this paper, we study the distribution of products of Pareto or Student’s <i>t</i> random variables that are ubiquitous in various risk factors analysis. We review known formulas for the probability density functions and derive their explicit forms for the products of Pareto and Gaussian or log-normal random variables. We also study how the Pareto or Student’s <i>t</i> random variable influences the asymptotic tail behaviour of the distribution of their product with the Gaussian or log-normal random variables and discuss how the dependency between the marginal random variables of the same type influences the probabilistic properties of the final product. The theoretical results are then applied for an analysis of the distribution of transaction values, being a product of prices and volumes, from a continuous trade on the German intraday electricity market.
Irene Gómez-Bueno, Sebastiano Boscarino, Manuel Jesús Castro
et al.
The aim of this work is to design implicit and semi-implicit high-order well-balanced finite-volume numerical methods for 1D systems of balance laws. The strategy introduced by two of the authors in a previous paper for explicit schemes based on the application of a well-balanced reconstruction operator has been applied. The well-balanced property is preserved when quadrature formulas are used to approximate the averages and the integral of the source term in the cells. Concerning the time evolution, this technique is combined with a time discretization method of type RK-IMEX or RK-implicit. The methodology will be applied to several systems of balance laws.