Abstract
The advantages of wind power integration over other renewable resources are well-known information and the natural results are the massive worldwide integration. Such massive integration, without the correct management together with conventional generation leads to an augmented complexity and the inflexibility of conventional power systems. For several reasons, forecasting tools are one of the most valuable tools in the power systems field, because they helps to decide in advance the way to manage correctly and with profits the electrical mix production. In this work, an extended hybrid wind power forecasting approach, with probabilistic features, is proposed to forecast twenty-four hours-ahead, considering only real historical wind power data. To validate the proposed forecasting approach, a comparison with other validated models is performed to offer a fair and proportional analysis. The outcomes show that the suggested forecasting approach performs adequately even considering the reduced data available as input.

Abstract
A viable option for electrification of remote areas far from power grids is to set up microgrids and feed them with local generation. Such microgrids are referred to as isolated microgrids and due to the lack of possibility of power exchange with the grid, their operation is different from grid-connected microgrids. Isolated microgrids, similar to grid-connected microgrids are equipped with energy management systems including unit commitment and economic dispatch modules. In this paper, the aim is to formulate the dynamic economic load dispatch (DELD) in isolated microgrids, while curtailment of responsive loads and curtailment of renewable power is allowed and load shedding is used as the last resort for balancing generation and demand. The generated power of dispatchable distributed generators (DGs), curtailed power of renewable DGs, curtailed demand and shed power are determined for each time period. The formulated DELD problem is solved with the well-established particle swarm optimisation (PSO) algorithm. The results for a microgrid with four dispatchable DGs and two renewable DGs show the performance of PSO over grey wolf optimisation (GWO) and also indicate the significant effect of demand response in reducing the operation cost of isolated microgrids.

Abstract

Abstract
Counting subgraph occurrences in complex networks is an important analytical task with applicability in a multitude of domains such as sociology, biology and medicine. This task is a fundamental primitive for concepts such as motifs and graphlet degree distributions. However, there is a lack of online algorithms for computing and updating subgraph counts in dynamic networks. Some of these networks exist as a streaming of edge additions and deletions that are registered as they occur in the real world. In this paper we introduce StreamFaSE, an efficient online algorithm for keeping track of exact subgraph counts in dynamic networks, and we explain in detail our approach, showcasing its general applicability in different network scenarios. We tested our method on a set of diverse real directed and undirected network streams, showing that we are always faster than the current existing methods for this task, achieving several orders of magnitude speedup when compared to a state-of-art baseline.

Abstract
O uso regular de jogos nas organizações requer integração destes nos processos e sistemas de gestão. Esta integração tem vindo a ser explorada a nível de investigação mas não é habitualmente trabalhada no ensino de desenvolvimento de videojogos. Pretende-se nesta apresentação resumir o conceito, apresentar potencialidades e expor formas de incluir o tema no ensino.

Abstract
Time series of (small) counts are common in practice and appear in a wide variety of fields. In the last three decades, several models that explicitly account for the discreteness of the data have been proposed in the literature. However, for multivariate time series of counts several difficulties arise and the literature is not so detailed. This work considers Bivariate INteger-valued Moving Average, BINMA, models based on the binomial thinning operation. The main probabilistic and statistical properties of BINMA models are studied. Two parametric cases are analysed, one with the cross-correlation generated through a Bivariate Poisson innovation process and another with a Bivariate Negative Binomial innovation process. Moreover, parameter estimation is carried out by the Generalized Method of Moments. The performance of the model is illustrated with synthetic data as well as with real datasets.