The article explores the effectiveness of Artificial Neural Networks (ANN) and XGBoost as surrogate models for building energy simulations, particularly when trained on limited data. It highlights the growing preference for ANNs in building performance simulation (BPS) but notes that XGBoost, a gradient-boosted regression tree model, often outperforms ANNs in terms of prediction accuracy and computational efficiency. The study also examines different sampling methods, finding that Monte Carlo-based sampling (MIPT) offers better prediction quality than Latin hypercube sampling (LHS). The research underscores the importance of model architecture, sample size, and input parameters in determining the accuracy of surrogate models, with XGBoost showing superior performance across various scenarios.

The study investigates the behavior of 'T' joints constructed from hollow steel sections using numerical simulations and experimental research. It highlights the importance of considering local deformations and stresses in the design of such connections, emphasizing the need for advanced numerical techniques like the finite element method (FEM) to accurately predict their ultimate bearing capacity. The research confirms that local plastification of the chord flange is a critical factor in the exhaustion of bearing capacity, and suggests that optimizing connection geometry can enhance load-bearing capacity. The findings support the need for updated European design regulations for 'T' connections.

The study investigates the prevalence of compensated advanced chronic liver disease (cACLD) and steatosis in patients with type 2 diabetes (T2D) using shear wave measurements (SWMs) and attenuation (ATT) measurements. It compares these methods with vibration-controlled transient elastography (VCTE) and the controlled attenuation parameter (CAP), which are established reference methods. The study found that among 214 T2D patients, 62.6% had steatosis and 8.9% had cACLD. SWMs correlated highly with VCTE for diagnosing cACLD, while ATT measurements were suboptimal for diagnosing liver steatosis. The study concludes that SWMs are a reliable diagnostic method for cACLD in T2D patients, but further research is needed to optimize ATT measurements for diagnosing liver steatosis.

A new simple experiment has been developed for students to quantitatively measure how the resistance of a wire depends on its temperature. Using minimal equipment such as a copper wire coil, a DC voltage source, a digital ohmmeter, and a calorimeter with an analog thermometer, students can discover a linear relationship between temperature and resistance. The temperature coefficient can be calculated from the graph's slope and compared with standard values. This experiment is suitable for high school or undergraduate physics labs and is of interest to teachers in introductory-level physics courses.

Oral health is a significant public health issue, particularly among Indigenous Australian children who experience higher rates of dental caries and related impacts compared to non-Indigenous children. The study validates the psychometric properties of the short forms of the Parental-Caregiver Perceptions Questionnaire (P-CPQ) and Family Impact Scale (FIS) for Indigenous children aged 2 to 3 years in South Australia. The resulting Aboriginal Children’s Oral Health-Related Quality of Life Questionnaire (A-COHQoL) is composed of three dimensions: Parent/Family Activities, COHQoL, and Family Conflict. The instrument demonstrated excellent psychometric properties and is recommended for future oral health studies and interventions among Indigenous children.