Abstract
The scarcity and price volatility of fossil fuels as well as environmental concerns has motivated the replacement of fossil fuel-powered vehicles by electric vehicles (EVs). Long charging time in battery charging stations is a serious barrier for large-scale adoption of EVs, so battery swap stations (BSSs) were developed wherein the near-empty batteries are exchanged with fully charged batteries and EV refilling is done in only a couple of minutes. Nowadays, BSSs are typically connected to a microgrid (MG) in their neighborhood. In this research, the optimal scheduling of MG resources and BSS is done for a grid-connected MG with dispatchable, photovoltaic and wind distributed generation (DG) units and operation cost of MG is minimised. It is assumed that the BSS services Tesla 3 EVs with 75 kWh batteries and a driving range of 496 km. A var compensator (VC) is connected to the MG that can purchase reactive power from var compensator. AC optimal power flow is done for the MG, while all network constraints, power loss and reactive power dispatch are taken into account and the cost of provision of reactive power is included in the operation cost of the MG. Generalized reduced-gradient (GRG) algorithm is used for the optimisation process. The effects of VC, optimal BSS scheduling and reactive power costs on active/ reactive power dispatch and MG operation cost are duly investigated.

Abstract
Mobile device users have been growing in the last years but the limited battery life of these devices is considered one of the major issues amongst users and programmers. Therefore, there is a need to guide developers in developing mobile applications in the most energy efficient way. One of the ways to improve this is to provide live feedback about the energy efficiency of a program while it's being programmed. We have analyzed and compared a total of 16 different tools and presented a list of 15 code smells and respective refactorings. From the analyzed tools, Leafactor is the closest to a valid solution to our problem because it's the only energy-aware tool with the highest liveness level. However, in order to be executed the programmer needs to trigger it on the IDE by selecting the file, instead of automatically being executed without the programmer being noticed and refactor his inefficient code.

Abstract
Multiple real-world problems are naturally modeled as cooperative multi-agent systems, ranging from satellite formation to traffic monitoring. These systems require algorithms that can learn successful policies with independent agents that rely solely on local partial-observations of the environment. However, multi-agent environments are more complex, due to their partial-observability and non-stationarity from an agent's perspective, as well as the structural credit assignment problem and the curse of dimensionality, and achieving coordination in such systems remains a complex challenge. To this end, we propose a multi-agent actor-critic algorithm called Asynchronous Advantage Actor Centralized-Critic with Communication (A3C3). A3C3 uses a centralized critic to estimate a value function, decentralized actors to approximate each agent's policy function, and decentralized communication networks for each agent to share relevant information with its team. The critic can incorporate additional information, like the environment's global state, when available, and optimizes the actor networks. The actor networks of an agent's teammates optimize that agent's communication network, such that each agent learns to output information that is relevant to the policies of others. A3C3 supports a dynamic amount of agents, noisy communication mediums, and can be horizontally scaled to shorten its learning phase. We evaluate A3C3 in two partially-observable multi-agent suites where agents benefit from communicating local information to each other. A3C3 outperforms state-of-the-art multi-agent algorithms, independent approaches, and centralized controllers with access to all agents' observations.

Abstract
Interactive storytelling uses in education are limited by the time required for its production and the ephemeral nature of interaction systems, leading inter-active stories to have a short usefulness life. We have developed the concept of platform-independent interactive stories, called virtual choreographies, enabling interactive stories to be replayed on novel technological platforms as they emerge, tackling the second half of this problem. The first part was also approached via a graphical storyboarding approach. Both aspects have been prototyped in a demonstration, called Magical Board Theatre (“Teatro de Tabuleiro Mágico”, original Portuguese name). We present this prototype, including its storyboarding tool, summarize the virtual choreographies ap-proach, and demonstrate how the prototype operationalizes them with story and platform examples.

Abstract
Machine Learning is a field in which significant steps forward have been taken in the last years, resulting in a wide variety of available algorithms, for many different problems. Nonetheless, most of these algorithms focus on the training of static models, in the sense that the model stops evolving after the training phase. This is increasingly becoming a limitation, especially in an era in which datasets are increasingly larger and may even arrive as sequential streams of data. Frequently retraining a model, in these scenarios, is not realistic. In this paper we propose evoRF: a combination of a Random Forest with an evolutionary approach. Its key innovative aspect is the evolution of the weights of the Random Forest over time, as new data arrives, thus making the forest's voting scheme adapt to the new data. Older trees can also be replaced by newly trained ones, according to their accuracy, ensuring that the ensemble remains up to date without requiring a whole retraining.

Abstract
Cardiovascular variability is the result of the activity of several physiological control mechanisms, which involve different variables and operate across multiple time scales encompassing short term dynamics and long range correlations. This study presents a new approach to assess the multiscale complexity of multivariate time series, based on linear parametric models incorporating autoregressive coefficients and fractional integration. The approach extends to the multivariate case recent works introducing a linear parametric representation of multiscale entropy, and is exploited to assess the complexity of cardiovascular and respiratory time series in healthy subjects studied during postural and mental stress.