137262 research outputs found
Sort by
Synthetic Resampling Algorithm for Response Class Imbalance in Supervised Learning: Application to Road Accident Severity Prediction
No full textRoad traffic injuries are a leading cause of death worldwide and are projected to become even more critical by 2030. Understanding the factors influencing crash severity is essential for developing effective safety interventions. However, crash data often suffer from severe class imbalance, especially when distinguishing between fatal and non-fatal accidents. Traditional machine learning algorithms tend to perform poorly under these conditions, favoring the majority class and misclassifying critical minority cases.
To address this, we propose a novel resampling algorithm—SONCA (Synthetic Over-sampling for Numerical and Categorical variables)—designed to balance datasets containing mixed data types. Unlike existing oversampling methods, SONCA handles numerical, ordinal, nominal, and dichotomous. We evaluated SONCA using both parametric (Logit) and non-parametric (CART) models on imbalanced datasets: PTW-ISTAT. The original models failed to detect the minority class effectively, while models estimated on SONCA-resampled data showed substantial improvements in True Positive Rate, G-mean, and Fmeasure. These results demonstrate SONCA’s potential as a flexible, model-agnostic preprocessing tool for addressing class imbalance in diverse real-world scenario
Post-pubertal Susceptibility to Cadmium-Induced Cardiac Injury: Role of Metallothionein, Antioxidant Defense, and Endocrine Modulation
No full textCadmium (Cd), a widespread environmental contaminant, induces cardiotoxicity even at low doses through endocrine disruption and oxidative stress. This investigation assessed age-related susceptibility in female albino rats by comparing pre-pubertal (30-day-old) and post-pubertal (60-day-old) animals that were administered CdCl2 at a dosage of 5.12 mg/kg body weight per day via oral gavage for 15 days. Despite greater Cd accumulation in pre-pubertal rats, post-pubertal animals exhibited heightened cardiac injury. This paradox correlated with impaired induction of metallothionein (MT), enhanced glutathione and vitamin C depletion, and reduced enzymic antioxidants (SOD, CAT, GPx). Concomitantly, altered adrenal function with increased corticosterone (CORT) and reduced estradiol (E2) reflected differential hypothalamic–pituitary–adrenal (HPA) axis responsiveness, further compromising antioxidant defenses. Biomarkers of cardiac function confirmed this vulnerability: increased serum CK-MB and troponin I, electrocardiographic (ECG) abnormalities (ST elevation, QT prolongation, tachycardia), and larger infarct volume. Collectively, these findings demonstrate that post-pubertal susceptibility is driven by reduced MT inducibility and HPA axis–mediated endocrine dysregulation, which amplify oxidative injury and structural remodeling of ventricular myocardium. From a translational standpoint, our findings identify adolescence and early adulthood as critical windows during which females exhibit heightened vulnerability to environmental CdCl2 exposure. Variations in metallothionein (MT) expression capacity and estradiol levels may serve as informative biomarkers for predicting individual susceptibility. These results also suggest the potential value of targeted interventions such as MT induction, antioxidant therapy, or modulation of hormonal pathways, that merit further investigation. Collectively, the study emphasizes the urgent need for stricter regulation of cadmium exposure, particularly among peripubertal populations, to reduce lifelong cardiovascular risk
Scavenger hunt: Selection of obscured active galactic nuclei combining multiband optical variability and colors
No full text
Sol-Gel Complements Conventional Strategies for the Synthesis of Self-Extinguishing Hybrid Silica-Epoxy Nanocomposites
No full textInquadramento normativo, requisiti procedurali e aspetti infrastrutturali delle decisioni amministrative automatizzate
No full textRelazione al Seminario «I dazi americani e il costituzionalismo “nel margine del caos”. La sentenza della Corte Suprema USA del 20 febbraio 2026», Dipartimento di Giurisprudenza, 16 febbraio 2026, ore 14.30, Aula A1.
No full text«I dazi americani e il costituzionalismo “nel margine del caos”. La sentenza della Corte Suprema USA del 20 febbraio 2026»
Methanol Sorption in Poly(ether imide): Molecular Insights from a Multiscale Study Combining Experiments, Theory, and Simulations
No full textThe objectives of this work are 2-fold. First, we aim to provide a comprehensive description of the poly(ether imide) (PEI)–methanol system, linking atomistic insights to macroscopic thermodynamic behavior, with explicit consideration of the hydrogen-bonding interactions governing the system. Second, by presenting this paradigmatic case, we intend to illustrate how a microscopic description validated through vibrational spectroscopy can be effectively exploited to establish a sound physical basis for developing a predictive macroscopic thermodynamic model. Sorption thermodynamics of methanol vapor in poly(ether imide) (Ultem 1000) has been thoroughly investigated through a synergistic multiscale theoretical approach combined with in situ FTIR spectroscopy. Density Functional Theory (DFT) calculations provided estimates of methanol–methanol and PEI–methanol interaction energies as well as simulated FTIR spectra. These results showed good agreement with the experimental outcomes of in situ FTIR spectroscopy, both in terms of the collected spectra and of the energies of hydrogen bond formation within the methanol-PEI system exposed to methanol vapor at different pressures and at a temperature of 30 °C. The analysis of the experimental FTIR spectra of the polymer phase enabled a quantitative assessment of the concentration of methanol molecules hydrogen-bonded to imide groups of the polymer and of methanol molecules hydrogen-bonded to the former methanol molecule (the so-called “first-shell” and ‘second-shell’ methanol molecules, respectively). The presence of these two “populations” was quantitatively confirmed by Molecular Dynamics (MD) simulations. The detailed picture emerging from these complementary approaches was then exploited to refine the Perturbed-Chain Statistical Associating Fluid Theory (PC-SAFT) equation of state model, which was used to interpret the sorption thermodynamics of methanol in PEI. In particular, the PC-SAFT model was extended, adopting the Dry Glass Reference Perturbation Theory (DGRPT), to address the nonequilibrium nature of the glassy PEI-methanol system and to account for the swelling associated with methanol sorption. Predictions made using the DGRPT-PC-SAFT theoretical framework showed a very good agreement with both the FTIR experimental results and the MD simulations, successfully predicting the concentrations of first- and second-shell methanol molecules in PEI as a function of methanol activity in the vapor phase
