Abstract
Futures contracts are a valuable market option for electricity negotiating players, as they enable reducing the risk associated to the day-ahead market volatility. The price defined in these contracts is, however, itself subject to a degree of uncertainty; thereby turning price forecasting models into attractive assets for the involved players. This paper proposes a model for futures contracts price forecasting, using artificial neural networks. The proposed model is based on the results of a data analysis using the spearman rank correlation coefficient. From this analysis, the most relevant variables to be considered in the training process are identified. Results show that the proposed model for monthly average electricity price forecast is able to achieve very low forecasting errors.

Abstract
Demand for leisure activities has increased due to some reasons such as increasing wealth, ageing populations and changing lifestyles, however, the efficiency of public transport system relies on solid demand levels and well-established mobility patterns and, so, providing quality public transportation is extremely expensive in low, variable and unpredictable demand scenarios, as it is the case of non-routine trips. Better prediction estimations about the trip purpose helps to anticipate the transport demand and consequently improve its planning. This paper addresses the contribution in comparing the traditional approach of considering municipality division to study such trips against a proposed approach based on clustering of dense concentration of services in the urban space. In our case, POIs (Points of Interest) collected from social networks (e.g. Foursquare) represent these services. These trips were associated with the territory using two different approaches: 'municipalities' and 'clusters' and then related with the likelihood of choosing a POI category (Points-of-Interest). The results obtained for both geographical approaches are then compared considering a multinomial model to check for differences in destination choice. The variables of distance travelled, travel time and whether the trip was made on a weekday or a weekend had a significant contribution in the choice of destination using municipalities approach. Using clusters approach, the results are similar but the accuracy is improved and due to more significant results to more categories of destinations, more conclusions can be drawn. These results lead us to believe that a cluster-based analysis using georeferenced data from social media can contribute significantly better than a territorial-based analysis to the study of non-routine mobility. We also contribute to the knowledge of patterns of this type of travel, a type of trips that is still poorly valued and difficult to study. Nevertheless, it would be worth a more extensive analysis, such as analysing more variables or even during a larger period.

Abstract
AppVox is a mobile application that provides support for children with speech and language impairments in their speech therapy sessions, while also allowing autonomous training at home. The application simulates a vocalizer with an audio stimulus feature, which can be used to train and amend the pronunciation of specific words through repetition. In this paper, we aim to present the development of the application as an assistive technology option, by adding new features to the vocalizer as well as assessing it as a usable option for daily training interaction for children with speech and language impairments. In this regard, we invited 15 children with speech and language impairments and 20 with no impairments to perform training activities with the application. Likewise, we asked three speech therapists and three usability experts to interact, assess, and give their feedback. In this assessment, we include the following parameters: successful conclusion of the training tasks (effectiveness); number of errors made, as well as number and type of difficulties found (efficiency); and the acceptance and level of comfort in completing the requested tasks (satisfaction). Overall, the results showed that children conclude the training tasks successfully and helped to improve their language and speech capabilities. Therapists and children gave positive feedback to the AppVox interface.

Abstract
Ovarian cancer is one of the pathologies with the worst prognostic in adult women and it has a very difficult early diagnosis. Clinical evaluation of gynaecological ultrasound images is performed visually, and it is dependent on the experience of the medical doctor. Besides the dependency on the specialists, the malignancy of specific types of ovarian tumors cannot be asserted until their surgical removal. This work explores the use of ultrasound data for the segmentation of the ovary and the ovarian follicles, using two different convolutional neural networks, a fully connected residual network and a U-Net, with a binary and multi-class approach. Five different types of ultrasound data (from beam-formed radio-frequency to brightness mode) were used as input. The best performance was obtained using B-mode, for both ovary and follicles segmentation. No significant differences were found between the two convolutional neural networks. The use of the multi-class approach was beneficial as it provided the model information on the spatial relation between follicles and the ovary. This study demonstrates the suitability of combining convolutional neural networks with beam-formed radio-frequency data and with brightness mode data for segmentation of ovarian structures. Future steps involve the processing of pathological data and investigation of biomarkers of pathological ovaries.

Abstract
This paper analyzes the electricity prices of the MIBEL electricity spot market with respect to a set of possible explanatory variables. Understanding the main drivers of the electricity price is a key aspect in understanding price formation and in developing forecasting models, which are essential for the selling and buying strategies of market agents. For this analysis, different techniques have been applied in this work, including standard and lasso regression models, causal analysis based on bayesian networks and classification trees. Results from the different approaches are coherent and show strong dependency of the electricity prices with the Portuguese imported coal for lower non-dispatchable net demands, which has been progressively replaced by gas for larger non-dispatchable net demands. Hydro reservoirs and hydro production are also main explanatory variables of the electricity price for all non-dispatchable net demand levels.

Abstract
Every year, thousands of new students begin their Masters in STEM related topics. Despite being regarded as a common occurrence by the faculty, it represents the culmination of years of studying and preparation for their professional life. Notwithstanding, these students face well-known recurrent problems: how to choose a topic, how to choose an advisor, how to start researching, and how to deal with all the unknowns associated with academic research. Although there are several books on how to write a thesis, most of them avoid prescriptive recommendations on topics beyond research per se or focus on doctoral students, for which the duration and motivation are significantly different. In this paper, we draft a pattern language comprised of thirty patterns that we have observed from supervising over a hundred masters students within the last decade.