Digital Elevation Models (DEM) are useful for solving various problems in civil works planning, risk assessment, hazard prediction, and spatial modeling. This research evaluated the accuracy of DEM generation produced from images obtained from unmanned aerial vehicles (UAV) in hovering mode (DEMf) and with eight Geodetic Control Points (DEMGCP), contrasted with the DEM generated from data from a Total Station - ST (DEMST) on the Bocanegra - Chachapoyas road, Peru. The photogrammetric images were processed with a spatial resolution of 50 cm. The mean, median, mode, mode, range, and standard deviation parameters were compared between DEMf, DEMGCP, and DEMST. In addition, statistical tests such as t-Student (α = 0.05) were applied for the difference of DEMs. The results showed a higher accuracy of the DEMGCP-derived vertical axis concerning DEMf with a combined error of 69.1 cm and an R2 of 0.999986; this is due to the use of GCPs to rectify the position on all axes of the generated DEM. Indeed, the use of UAVs is of great applicability in the generation of DEMs. They complement the data from a total station for the formulation, development, and implementation of civil works projects.
This study evaluated the technological and sensory effects of micronized salt and yeast extract as sodium-reduction strategies in beef burgers. Five formulations were developed: a control (C) with 1.75% regular salt, and four formulations with 50% salt reduction: RR (0.875% regular salt), RM (0.875% micronized salt), RM1 (0.875% micronized salt +1% yeast extract), and RM2 (0.875% micronized salt +2% yeast extract). Salt reduction, the use of micronized salt, and yeast extract incorporation did not significantly affect color, pH, fat content, diameter reduction, fat retention, and some texture parameters such as springiness, cohesiveness, and chewiness of the burgers compared with the control. Reduced-salt burgers showed higher water activity and lower ash and sodium contents than the control, with sodium levels ranging from 650 mg/100 g (RM1) to 1073.24 mg/100 g (C). They also showed higher cooking loss and lower moisture retention, while yeast extract addition increased tenderness. Lipid oxidation did not differ among raw samples; however, in cooked burgers, RM1 and RM2 exhibited significantly higher TBARS values, which remained within acceptable sensory thresholds. In the sensory evaluation, burgers with 50% salt reduction, micronized salt, and 2% yeast extract were perceived as “just right in salt” and achieved overall liking scores comparable to those of the control. These results demonstrate that the combination of micronized salt and yeast extract is an effective strategy for reducing sodium content in beef burgers without compromising sensory acceptance, offering a viable approach for the development of healthier meat products.
This study analyzes the relationship between frequency indices (FI), severity indices (SI), accident rates (AR), hazardous incidents, and fatal accidents in Peruvian mining (2010–2023). It addresses the issue of methodological differences in calculating these indices across sectors and their influence on preventing fatal accidents. Using data from MINEM, a descriptive-explanatory approach was applied, analyzing statistics on fatal accidents and hazardous incidents through descriptive and inferential statistical methods. The results reveal a very weak correlation (0.253) between hazardous incidents and fatal accidents, a moderate positive correlation between FI and fatal accidents (0.507–0.633), and a strong correlation (0.808) between SI and fatal accidents, highlighting that severity has a more direct impact on fatalities. Finally, AR shows a moderate positive correlation (0.672) with fatal accidents. It concludes that safety indicators are related to the number of fatal accidents; however, preventing fatal accidents requires comprehensive strategies focused on reducing the frequency and severity of accidents, beyond merely monitoring indices.
BackgroundThe rapid expansion of artificial intelligence tools in higher education demands understanding whether students' moral development is associated with more responsible AI use. Moral development, conceptualized through Kohlberg's stage theory, provides a theoretically grounded framework for examining this relationship, yet few empirical studies have tested it in a Latin American university context.MethodsA quantitative, cross-sectional, correlational design was used with a quota sample of 5,487 Peruvian university students recruited through institutional channels. Moral development was assessed using the Defining Issues Test-2 (DIT-2) and responsible AI use was measured with the EURIA-ES, a 24-item self-report scale with evidence of reliability and factorial structure comprising four dimensions: transparency, academic honesty, critical algorithmic awareness, and responsibility in content dissemination.ResultsMost participants fell in the middle descriptive N2 band (61.4%; N2 mean = 32.14), using descriptive cut points rather than validated Kohlbergian stage diagnoses. EURIA-ES scores were moderate (M = 72.34 out of 120), with academic honesty scoring highest and critical algorithmic awareness lowest. The N2 index showed an important standardized association with EURIA-ES scores (beta = .31). The final model accounted for 23.5% of the variance (R2 = .235), and entering N2 increased explained variance by 9.2 percentage points beyond the sociodemographic block (ΔR2 = .092). Interpretation emphasizes effect magnitude and 95% confidence intervals rather than statistical significance alone. Year of study and field of knowledge showed adjusted associations, whereas regional differences were small in magnitude.Conclusionswithin a cross-sectional design, higher DIT-2 N2 scores were associated with higher self-reported responsible academic use of generative AI. N2 should be interpreted as an important correlate rather than a primary determinant of responsible AI use. These findings pertain specifically to academic uses of generative AI and should not be generalized to all artificial intelligence systems or to objectively observed student behavior.
Dengue case forecasting is important for the prevention and early control of outbreaks, as well as for the optimization of healthcare resources, among other aspects. This study addresses the need to develop increasingly accurate forecasting models that can support informed decision-making before and during dengue epidemics. Accordingly, two new models based on convolutional and recurrent neural networks, namely ConvLSTM and ConvBiLSTM, combined with data augmentation based on linear interpolation, are proposed. As a case study, weekly dengue cases in Peru from 2000 to 2024 are used. The proposed models are compared with well-known recurrent neural network-based models such as LSTM, BiLSTM, GRU, and BiGRU, both with and without data augmentation. The results show that the proposed models with data augmentation achieve comparable and superior performance to the benchmark models, while also exhibiting a lower average computational cost.
