Abstract
Collaborative Filtering (CF) has become the standard approach to solve recommendation systems problems. Collaborative Filtering algorithms try to make predictions about interests of a user by collecting the personal interests from multiple users. There are multiple CF algorithms, each one of them with its own biases. It is the Machine Learning practitioner that has to choose the best algorithm for each task beforehand. In Recommender Systems, different algorithms have different performance for different users within the same dataset. Meta Learning has been used to choose the best algorithm for a given problem. Meta Learning is usually applied to select algorithms for a whole dataset. Adapting it to select the to the algorithm for a single user in a RS involves several challenges. The most important is the design of the metafeatures which, in typical meta learning, characterize datasets while here, they must characterize a single user. This work presents a new meta-learning based framework named µ-cf2vec to select the best algorithm for each user. We propose using Representation Learning techniques to extract the metafeatures. Representation Learning tries to extract representations that can be reused in other learning tasks. In this work we also implement the framework using different RL techniques to evaluate which one can be more useful to solve this task. In the meta level, the meta learning model will use the metafeatures to extract knowledge that will be used to predict the best algorithm for each user. We evaluated an implementation of this framework using MovieLens 20M dataset. Our implementation achieved consistent gains in the meta level, however, in the base level we only achieved marginal gains. © 2020 IEEE.

Abstract
The growing digitalisation, grid complexity and the number of digitally connected devices that communicate with systems in the distribution grid are enabling the continuous development of automation and intelligence, acquisition of data from sensors, meters and devices for monitoring and managing the distribution network, to achieve an enhanced, preventive, resilient and flexible network operation philosophy. This study presents a set of use cases towards the demonstration of the benefits of implementing a platform that collects, aggregates and facilitates horizontal integration and data correlation from various sources, enabling these use cases across the distribution grid. The adequacy analysis of current distribution network architecture considered derived requirements on the characterisation of its evolution taking advantage of key digital technologies, towards the implementation of distributed control and management strategies. It is also presented a benefit analysis of implementing a centralised common data and service platform, i.e. an internet of things (IoT) platform, regarding new functionalities and applications.

Abstract

Abstract
Internationalization at Home (IaH) is the most accessible approach for internationalizing education, as it does not involve mobility or considerable investment. This article discusses results of two distinct IaH initiatives: a 4-week collaboration between students from a Portuguese university and a Mexican university, and a set of activities conducted throughout one semester in a multicultural classroom in one Portuguese university. The analysis shows that, despite the clear differences of the two initia-tives, they provided very interesting outcomes, with students recogniz-ing the development of intercultural communication skills and other soft skills, which were perceived as adding value to the learning process and to their future professional careers.

Abstract
Background: Disease-related malnutrition is a significant problem in hospitalized patients, with high prevalence rates depending on the studied population. Internal Medicine wards are the backbone of the hospital setting. However, prevalence and determinants of malnutrition in these patients remain unclear. We aimed to determine the prevalence of malnutrition in Internal Medicine wards and to identify and characterize malnourished patients. Methods: A cross-sectional observational multicentre study was performed in Internal Medicine wards of 24 Portuguese hospitals during 2017. Demographics, hospital admissions during the previous year, type of admission, primary diagnosis, Charlson comorbidity index, and education level were registered. Malnutrition at admission was assessed using Patient-Generated Subjective Global Assessment (PG-SGA). Demographic characteristics were compared between well-nourished and malnourished patients. Logistic regression analysis was used to identify determinants of malnutrition. Results: 729 participants were included (mean age 74 years, 51% male). Main reason for admission was respiratory disease (32%). Mean Charlson comorbidity index was 5.8 +/- 2.8. Prevalence of malnutrition was 73% (56% moderate/suspected malnutrition and 17% severe malnutrition), and 54% had a critical need for multidisciplinary intervention (PG-SGA score >= 9). No education (odds ratio [OR] 1.88, 95% confidence interval [CI]: 1.16-3.04), hospital admissions during previous year (OR 1.53, 95%CI: 1.05-2.26), and multiple comorbidities (OR 1.22, 95%CI: 1.14-1.32) significantly increased the odds of being malnourished. Conclusions: Prevalence of malnutrition in the Internal Medicine population is very high, with the majority of patients having critical need for multidisciplinary intervention. Low education level, admissions during previous year, and multiple comorbidities increase the odds of being malnourished.

Abstract
The new 5G mobile communication system era brings a new set of communication devices that will appear on the market. These devices will generate data streams that require proper handling by machine algorithms. The processing of these data streams requires the design, development, and adaptation of appropriate machine learning algorithms. While stream processing algorithms include hyper-parameters for performance refinement, their tuning process is time-consuming and typically requires an expert to do the task. In this paper, we present an extension of the Self Parameter Tuning (SPT) optimization algorithm for data streams. We apply the Nelder-Mead algorithm to dynamically sized samples that converge to optimal settings in a double pass over data (during the exploration phase), using a relatively small number of data points. Additionally, the SPT automatically readjusts hyper-parameters when concept drift occurs. We did a set of experiments with well-known classification data sets and the results show that the proposed algorithm can outperform the results of previous hyper-parameter tuning efforts by human experts. The statistical results show that this extension is faster in terms of convergence and presents at least similar accuracy results when compared with the standard optimization techniques. © 2020, Springer Nature Switzerland AG.