Insulin sensitivity prediction is crucial for model-based treatment in Intensive Care Unit patients, particularly those with hyperglycemia. However, predicting insulin sensitivity is challenging due to inter- and intra-patient variability. Different neural network models are proposed and compared for predicting insulin sensitivity, including recurrent and feedforward versions of the Classification Deep Neural Network and Mixture Density Network models. These models were trained using 1879 patient records containing 123,988 insulin sensitivity values from three intensive care patient cohorts in three different countries. Results show that using patient history in prediction models can improve the accuracy of insulin sensitivity predictions. The Mixture Density Network model provided more accurate predictions, measured by a problem-specific metric that expresses clinical requirements. We demonstrated that even using up to 12 h of historical data can improve prediction accuracy. This study highlights the potential of recurrent neural network models in predicting insulin sensitivity in Intensive Care Unit patients. Our findings suggest that using recurrent neural networks and incorporating patient history can lead to more accurate predictions. These results are generalizable due to the large and diverse dataset employed, which included patients from three different cohorts in three care settings.

Neural circuitry remodeling in chronic pain and depression comorbidity: Toward an emotion-perception integration network model.

Core insomnia symptoms associated with cognitive flexibility in insomnia disorder: A network analysis.

A coupling network for brain computing: E-I balanced embedding in dual-attractor dynamics systems.

Stability control of metastable states as a unified mechanism for flexible temporal modulation in cognitive processing.

• Weighted ensembles were developed for multi-day allergenic pollen forecasting. • The approach showed a robust performance and spatial transferability to new locations. • Models showed promise for regions with sparse monitoring using non‐local pollen data. • The impact of weather and phenological indicators on pollen levels was assessed. • Methodology provides a solution for allergy management in data-scarce environments. Airborne allergenic birch (Betula) and grass (Poaceae) pollen poses significant public health challenges, as high concentrations can trigger allergic rhinitis and exacerbate asthma for millions of people globally. Accurate forecasts of pollen concentrations enable vulnerable individuals to take preventive measures and support healthcare planning. In this study, we developed a spatially transferable forecasting methodology based on weighted ensemble machine learning models to predict daily birch and grass pollen concentrations up to five days ahead. Historical daily pollen measurements (2006–2022) from a Hirst-type trap in Wrocław, Poland, were combined with Weather Research & Forecasting (WRF) meteorological outputs (e.g., temperature, humidity, wind speed, precipitation), phenological indices (e.g., growing degree days), calendar features (e.g., day of year, week), and lagged pollen predictors. Four decision tree-based base learners – Random Forest, ExtraTrees, XGBoost, and LightGBM – were tuned within nested leave‐year‐out cross-validation and combined via a weighted‐average ensemble. Separate ensembles were trained for lead times from +1 to +5 days for both taxa. Variable importance analysis revealed that lagged pollen predictors dominated short‐term forecasts, while phenological and meteorological variables gained prominence at longer lead times. Independent testing on 2023–2024 data covered Wrocław and three other Polish cities, with two scenarios: (A) using local lagged pollen inputs and (B) replacing local inputs with non-local data from Wrocław (applied only to external sites). Successful spatial transferability was demonstrated in both scenarios. Ensemble models consistently outperformed individual algorithms, achieving test coefficients of determination (R²) of ∼0.77 (birch) and ∼0.72 (grass) at +1 day, declining to ∼0.55 and ∼0.66 at +5 days (scenario A).The findings illustrate that a weighted ensemble can produce reliable short‐ and medium‐term pollen forecasts in data‐scarce regions even if only a single well‐monitored site is available. The approach is readily adaptable for operational implementation and may improve allergy prevention and public‐health interventions in regions with limited pollen monitoring.

Heat transfer characteristics of nonlinearly graded metal foam in thermal storage tank: A novel framework model of BP neural network fused with tactical unit algorithm

Study on methods for measuring beef color and predicting storage time based on computer vision.

Modeling the empathy-self-discovery paradox in Gen Z social behavior with NPD using artificial neural networks.

Hybrid convolutional neural network and selective state space model with integrated edge features for infrared small target detection

Investigation of self-pressurization in liquid hydrogen storage tanks using a novel coupled multi-node non-equilibrium thermodynamic and two-dimensional thermal resistance network model

Improved AHP and dual-hidden layer adaptive regularized neural network model for the quantitative risk assessment of oil and gas pipelines

Tabular foundation models as a new portable standard in local surgical risk prediction.

Response of microbial nitrogen removal to sinuosity in river bends: mechanisms and development of physics-informed neural networks model.

Assamese glutinous Bora rice (Oryza sativa L.) is widely used for various ethnic food preparations. However, its resistant starch (RS) content, which influences the glycemic index (GI), remains poorly characterized. This exploratory study examined nine popular cooking and eating quality (CEQ) traits in 21 Bora rice lines, and performed molecular characterization and expression profiling during grain development, emphasizing machine learning (ML)-based prediction of RS content. The endosperm of Bora rice lines contains 80% to 90% starch, predominantly amylopectin, with a lower proportion of RS. Low gelatinization temperature, shorter cooking times at boiling temperatures, and soft gel length are key physicochemical traits of this group. Oryza sativa L. 'Aghani Bora' requires 68 minutes to prepare fully at room temperature. This reflects its low gelatinization temperature and soft gel formation, which are characteristic of Bora rice. Glycemic index-linked polymorphic markers can support molecular breeding of Bora rice for low GI. GBSSI and SSIIa transcripts were downregulated in genotypes exhibiting low RS content. Significant correlations were observed among CEQ traits. The radial basis function network model for predicting RS content in Bora rice yielded a high R2 (0.9155) and a low mean squared error (0.0690). Amylose appears to have a critical role in determining most CEQ characteristics but has less influence on readiness to eat. Bora rice requires genetic improvement because its low RS content can lead to a high GI. The low-cost machine learning (ML) model developed in this study provides an effective tool for rapid prediction of RS content in rice and other starchy cereal crops. © 2026 Society of Chemical Industry.

Spectrum governance and 5G market structure: Malaysia's transition to a dual network model

GeoLSTM: A road geometry-aware neural network model for spatio-temporal highway traffic speed and intensity forecasting

Water-energy-food nexus in arid and semi-arid regions under spatial complexity: A case study in Ningxia, China.

Bayesian network model for assessing hydrogen ignition probability

Predicting the tabletability of binary mixtures from individual powder compaction behavior.