Abstract
The paper presents the comparison of accuracy in the Speech Emotion Recognition task using the Hamming and Hanning windows for framing the speech and determining the spectrogram to be used as input of a convolutional neural network. The detection of between 4 and 10 emotional states was tested for both windows. The results show significant differences in accuracy between the two window types and provide valuable insights for the development of more efficient emotional state detection systems. The best accuracy between 4 and 10 emotions was 64.1% (4 emotions), 57.8% (5 emotions), 59.8% (6 emotions), 48.4% (7 emotions), 47.8% (8 emotions), 51.4% (9 emotions), and 45.9% (10 emotions). These accuracy is at the state-of-the art level.

Abstract
The transition to college is a period of higher risk of the development of eating disorders, with nutrition/dietetics students representing a group of particular vulnerability. Hence, it is interesting to assess eating disorders, taking into consideration potential sources of bias, including social desirability. Our aims were to compare the risk of eating disorders between students of nutrition/dietetics and those attending other courses and to study potential social desirability biases. A total of 799 higher education students (81.7% females) aged 18 to 27 years old completed a questionnaire assessing the risk of eating disorders (EAT-26) and social desirability (composite version of the Marlowe-Crowne Social Desirability Scale). The proportion of students with a high risk of eating disorders was higher among females (14.5% vs. 8.2%, p = 0.044). Nutrition/dietetics students did not differ from those attending other courses regarding the risk of eating disorders. The social desirability bias when assessing the risk of eating disorders was overall low (EAT-26 total score: r = -0.080, p = 0.024). Social desirability correlated negatively with the Diet (r = -0.129, p < 0.001) and Bulimia and food preoccupation subscales (r = -0.180, p < 0.001) and positively with Oral self-control (r = 0.139, p < 0.001).

Abstract
During the process of software architectural design, numerous questions arise which must be answered. These questions may be about requirements on the proposed system (the problem space) or about how the system should be designed and developed (the solution space). As questions arise they may be answered immediately, deferred until later, or provisionally answered with an assumption about the answer. The objective of this work was to explore the nature of questions that arise during architecture. We explored the types of questions, how they are organized, how they are tracked, and how and when they are answered. We started by surveying highly experienced architects about their practices with respect to architectural questions. We also performed a controlled experiment with master students about organizing architectural questions that clarified and substantiated the survey data. We learned that architectural questions include slightly more questions about the problem space than the solution space, as well as a minority of questions related to the managing of the project. We found that architects often use ad hoc methods to organize and track them, although they typically organize them along more than one dimension. We learned also that, about a third of the time, architects make assumptions about the answers to architectural questions in order to make progress on the architecture. This suggests that some projects may have risks of incorrect design or later costly rework due to inadequate tracking or incorrectly answered architectural questions.

Abstract

Abstract
There is evidence that some outdoor events may have contributed to the spread of COVID-19. We updated an empirical methodology based on regression modeling and hypothesis testing to analyze the potential impact of a demonstration that took place in Lisbon, within the scope of the ’Black Lives Matter’ context, on the contagion pattern in the region where this event occurred. We find that in the post-impact period there was no acceleration in the number of cases in the region, unlike in a prior event in the region. The proportion of counties where there was a potential impact of the event is not statistically significant. This result demonstrates that not all outdoor events contributed to the spread of COVID-19 and exemplifies how to apply the selected empirical methodology.

Abstract
This paper introduces a novel approach to autonomous vehicle control using an end-to-end learning framework. While existing solutions in the field often rely on computationally expensive architectures, our proposed lightweight model achieves comparable efficiency. We leveraged the Car Learning to Act (CARLA) simulator to generate training data by recording sensor inputs and corresponding control actions during simulated driving. The Mean Squared Error (MSE) loss function served as a performance metric during model training. Our end-to-end learning architecture demonstrates promising results in predicting steering angle and throttle, offering a practical and accessible solution for autonomous driving. Results of the experiment showed that our proposed network is approximate to 5.4 times lighter than Nvidia's PilotNet and had a slightly lower testing loss. We showed that our network is offering a balance between performance and computational efficiency. By eliminating the need for handcrafted feature engineering, our approach simplifies the control process and reduces computational demands. Experimental evaluation on a testing map showcases the model's effectiveness in real-world scenarios whilst being competitive with other existing models.