Abstract
The grape variety plays an important role in the wine production chain, thus identifying it is crucial for production control. Ampelographers, professionals who identify grape varieties through plant visual analysis, are scarce, and molecular markers are expansive to identify grape varieties on a large scale. In this context, Deep Learning models become an effective way to handle ampelographers scarcity. In this work, we explore the benefit of using deep learning vision transformers architecture relative to conventional CNN to identify 12 grapevine varieties using leaf-centred RGB images acquired in the field. We train an Xception model as a baseline and four different configurations of the ViT_B model. The best model achieved 0.96 of F1-score, outperforming the state-of-the-art convolutional-based model in the used dataset.

Abstract
The use of deep learning methods in medical imaging has been able to deliver promising results; however, the success of such models highly relies on large, properly annotated datasets. The annotation of medical images is a laborious, expensive, and time-consuming process. This difficulty is increased for the mutations status label since these require additional exams (usually biopsies) to be obtained. On the other hand, raw images, without annotations, are extensively collected as part of the clinical routine. This work investigated methods that could mitigate the labelled data scarcity problem by using both labelled and unlabelled data to improve the efficiency of predictive models. A semi-supervised learning (SSL) approach was developed to predict epidermal growth factor receptor (EGFR) mutation status in lung cancer in a less invasive manner using 3D CT scans.The proposed approach consists of combining a variational autoencoder (VAE) and exploiting the power of adversarial training, intending that the features extracted from unlabelled data to discriminate images can help in the classification task. To incorporate labelled and unlabelled images, adversarial training was used, extending a traditional variational autoencoder. With the developed method, a mean AUC of 0.701 was achieved with the best-performing model, with only 14% of the training data being labelled. This SSL approach improved the discrimination ability by nearly 7 percentage points over a fully supervised model developed with the same amount of labelled data, confirming the advantage of using such methods when few annotated examples are available.

Abstract
Carlos Baquero and Rosa Cabecinhas consider how readers make assumptions about authors’ roles and relative contributions when reading papers. It is not unexpected that when reading papers, readers also make simplifications and have assumptions about author roles and relative contributions. Experts also observed that the success of a new work depends not only on its factual quality, but on the prior recognition of the author and its institution. Work done at more prestigious departments can diffuse more rapidly through the science networks. The bias that occurs both on author and institution recognition is now well-known and a justification for blind review mechanisms.

Abstract
Este artigo apresenta e discute uma prática pedagógica cocriada no contexto do componente curricular Seminário Temático “Educação & Transformação Digital”, um dos produtos que tem origem na relação entre duas pesquisas desenvolvidas em nível de internacionalização: “Transformação Digital e Humanidades: Educação e Comunicação em Movimento”(CAPES PrInt) e “O Urbano e o Pós-Urbano como Espaços de Aprendizagem: formação de professores e pesquisadores na cultura híbrida e multimodal” (FAPERGS). A Atividade Acadêmica é ofertada, conjuntamente, na modalidade on-line, para mestrandos e doutorandos dos programas de Pós-Graduação da Unisinos e da Universidade Aberta de Portugal (UAb). O referido seminário visou problematizar os desafios e cenários emergentes na área da Educação Digital e, também, como as diferentes instituições estão se apropriando das transformações digitais a fim de provocar inovações na Educação. A pesquisa é de natureza exploratória e descritiva, de abordagem qualitativa, e se apropria do Método Cartográfico de Pesquisa-Intervenção, para produção e análise de dados. Nesse sentido, como um dos resultados do percurso investigativo e problematizador realizado no Seminário, emergiu a vivência “Novas Aventuras de Dom Quixote”, uma prática pedagógica inventiva. Esta vivência, fundamentada nas Epistemologias Reticulares e Conectivas, foi cocriada na perspectiva de uma Educação online/onlife, objetivando a apropriação de diferentes Tecnologias Digitais (TD), no enfrentamento de desafios contemporâneos de aprendizagem e que buscou instigar a apropriação das mesmas, no sentido de compreendê-las enquanto potência, que pode contribuir para propiciar vivências de aprendizagem e a ampliação dos espaços educacionais, uma vez que possibilitam a transubstanciação da educação, do espaço físico das instituições educacionais, para o formato on-line.

Abstract
Determining the transformer top-oil temperature (TOT) is one of the key issues in determining the transformer insulation life and reliability of the power system. Due to the non-linear nature of the model presented in the IEEE C57.91 standard to determine this temperature, a more precise method is needed to estimate the equation coefficients to estimate the TOT in the future. This paper presents a method for online thermal modelling of the transformer according to the IEEE C57.91 based on the Unscented Kalman filter (UKF). This method can be applied to transformers with a variety of cooling modes and estimates the TOT with an acceptable error. In order to evaluate the proposed method, the practical data of the 800 kVA distribution transformer with unknown equation coefficients and simulated data with known coefficients are used, and finally, by calculating the estimation error, the proper performance of the presented method is proved. It is proved that the proposed method predicts TOT even in the presence of noise with an error of less than 0.5 degrees C and a delay of less than 1.5 h. It makes the proposed method can be implemented for purposes such as load management, and insulation life estimation of the transformer.

Abstract
The interconnection of AC and DC microgrids results in a hybrid AC/DC microgrid (HMG). In light of HMGs, the future smart grid implementation will be facilitated. One important aspect in HMGs is the interconnection of AC and DC microgrids and control of bidirectional interlink power converters (BILPCs), which has taken a lot of research attention in the last decade. The BILPCs are the most prevalent method for interconnection of HMGs. Thus, the current study first reviews different interconnection methods and control challenges of AC and DC microgrids in HMGs and then overviews various control strategies of BILPCs presented in literature, all carried out in a comprehensive manner.