Abstract
A pandemia Covid-19 teve um profundo impacto nos processos pedagógicos das universidades de ensino presencial. O confinamento obrigatório da população universitária implicou a transposição do ensino presencial em sala para as aulas realizadas online, levando a um aumento considerável do uso de sistemas de videoconferência, estabelecendo novas (ou reforçando as existentes) comunidades de aprendizagem online. Estas práticas de aprendizagem online tendem a se manter após o fim da pandemia na medida em que proporcionem alternativas complementares de contacto, partilha e colaboração em rede, determinando formas não estruturadas de modelos pedagógicos híbridos de ensino a distância. Neste artigo descrevemos práticas concretas de ensino online implementadas durante a pandemia em duas universidades de ensino presencial, a Universidade de São José em Macau; e a Universidade de São Paulo, Brasil, procedendo à sua análise crítica à luz da práxis de ensino a distância online em voga na Universidade Aberta, Portugal, uma universidade virtual de ensino a distância, tendo em vista a definição de um modelo pedagógico virtual de cariz geral que possa proporcionar princípios e linhas mestras para o planeamento, organização e implementação de oferta educativa de nível universitário administrável online para os tempos pós-pandémicos suportada nas evidências da pesquisa e tendências emergentes. 

Abstract
Cognitive Digital Twin (CDT) has been taking considerable attention in several recent studies. CDT is considered as a promising evolution of Digital Twin bringing new smart and cognitive capabilities. Therefore, it is important to understand how companies can exploit this new technology and create new data-driven business models. Considering that context this article aims to identify Smart PSS business model based on Cognitive Digital Twin platforms. To reach this goal a literature review was conducted. As a principal contribution this study brings a set of new business models to offer Smart PSS based on cognitive digital twins. Moreover, the study presents several real cases of companies that are currently using the cognitive capabilities supplied by edge companies of the digital twin technologies. © 2022, IFIP International Federation for Information Processing.

Abstract
The use of several distributed generators as well as the energy storage system in a local microgrid require an energy management system to maximize system efficiency, by managing generation and loads. The main purpose of this work is to find the optimal set-points of distributed generators and storage devices of a microgrid, minimizing simultaneously the energy costs and the greenhouse gas emissions. A multi-objective approach called Pareto-search Algorithm based on direct multi-search is proposed to ensure optimal management of the microgrid. According to the non-dominated resulting points, several scenarios are proposed and compared. The effectiveness of the algorithm is validated, giving a compromised choice between two criteria: energy cost and GHG emissions.

Abstract
In this paper, we consider a variant of adaptive robust combinatorial optimization problems where the decision maker can prepare K solutions and choose the best among them upon knowledge of the true data realizations. We suppose that the uncertainty may affect the objective and the constraints through functions that are not necessarily linear. We propose a new exact algorithm for solving these problems when the feasible set of the nominal optimization problem does not contain too many good solutions. Our algorithm enumerates these good solutions, generates dynamically a set of scenarios from the uncertainty set, and assigns the solutions to the generated scenarios using a vertex p-center formulation, solved by a binary search algorithm. Our numerical results on adaptive shortest path and knapsack with conflicts problems show that our algorithm compares favorably with the methods proposed in the literature. We additionally propose a heuristic extension of our method to handle problems where it is prohibitive to enumerate all good solutions. This heuristic is shown to provide good solutions within a reasonable solution time limit on the adaptive knapsack with conflicts problem. Finally, we illustrate how our approach handles nonlinear functions on an all-or-nothing subset problem taken from the literature.

Abstract
It has been known for a while that the problem of multi-source domain adaptation can be regarded as a single source domain adaptation task where the source domain corresponds to a mixture of the original source domains. Nonetheless, how to adjust the mixture distribution weights remains an open question. Moreover, most existing work on this topic focuses only on minimizing the error on the source domains and achieving domain-invariant representations, which is insufficient to ensure low error on the target domain. In this work, we present a novel framework that addresses both problems and beats the current state of the art by using a mildly optimistic objective function and consistency regularization on the target samples.

Abstract