Abstract

Abstract
This paper aims to understand the importance of ICT (Information and Communication Technologies) in the tourism sector, focusing on the Island of Porto Santo. We are in an increasingly digital world and information and communication technologies represents a way to be up to date and present in the market. The use of social networks by companies, are used as means of communication and promotion, with the purpose of understanding the behavior of tourist consumers. In order to understand the importance of ICT and social networks in the tourism sector of the Island of Porto Santo, a questionnaire was prepared and distributed in several places on the Island of Porto Santo. The results are presented in this article. For a better understanding of ICT and social networks, a literature review on the topic was also carried out.

Abstract
In this paper, the RT-WiFi architecture is proposed to handle real-time (RT) communication in infrastructured IEEE 802.11 networks operating in high density industrial environments. This architecture is composed of a time division multiple access (TDMA)-based coordination layer that schedules the medium access of RT traffic flows, and an underlying traffic separation mechanism that is able do handle the coexistence of RT and non-RT traffic sources in the same communication environment. The simulation assessment considers an overlapping basic service set (OBSS), where a set of RT and non-RT stations share the same frequency band. The performance assessment compares the behaviour of the RT-WiFi architecture vs. the behaviour of standard distributed coordination function (DCF), point coordination function (PCF), enhanced distributed channel access (EDCA), and hybrid coordination function (HCF) controlled channel access (HCCA) medium access control mechanisms. A realistic error-prone model has been used to measure the impact of message losses in the RT-WiFi architecture. It is shown that the proposed RT-WiFi architecture offers a significantly enhanced behaviour when compared with the use of IEEE 802.11 standard mechanisms, in what concerns average deadline misses and average access delays. Moreover, it also offers an almost constant access delay, which is a relevant characteristic when supporting RT applications.

Abstract
The small amount of public available medical images hinders the use of deep learning techniques for mammogram automatic diagnosis. Deep learning methods require large annotated training sets to be effective, however medical datasets are costly to obtain and suffer from large variability. In this work, a lightweight deep learning pipeline to detect, segment and classify anomalies in mammogram images is presented. First, data augmentation using the ground-truth annotation is performed and used by a cascade segmentation and classification methods. Results are obtained using the INbreast public database in the context of lesion detection and BI-RADS classification. Moreover, a pre-trained Convolutional Neural Network using ResNet50 is modified to generate the lesion regions proposals followed by a false positive reduction and contour refinement stages while a pre-trained VGG16 network is fine-tuned to classify mammograms. The detection and segmentation stage results show that the cascade configuration achieves a DICE of 0.83 without massive training while the multi-class classification exhibits an MAE of 0.58 with data augmentation.

Abstract
This paper addresses a distribution problem involving a set of different products that need to be distributed among a set of geographically disperse retailers and transported from the single warehouse to the aforementioned retailers. The disfribution and transportation are made in order to satisfy retailers' demand while satisfying storage limits at both the warehouse and the retailers, transportation limits between the warehouse and the retailers, and other operational constraints. This problem is combinatorial in nature as it involves the assignment of a discrete finite set of objects, while satisfying a given set of conditions. Hence, we propose a genetic algorithm that is capable of finding good quality solutions. The genetic algorithm proposed is used to a real case study involving the disfribution of eight products among 108 retailers from a single warehouse. The results obtained improve on those of company's current practice by achieving a cost reduction of about 13%.

Abstract
Challenges in the coordination between the transmission system operator (TSO) and the distribution system operator (DSO) have risen continuously with the integration of distributed energy resources (DER). These technologies have the possibility to provide reactive power support for system operators. Considering the Portuguese reactive power policy as an example of the regulatory framework, this paper proposes a methodology for proactive reactive power management of the DSO using the renewable energy sources (RES) considering forecast uncertainty available in the distribution system. The proposed method applies a stochastic sequential alternative current (AC)-optimal power flow (SOPF) that returns trustworthy solutions for the DSO and optimizes the use of reactive power between the DSO and DER. The method is validated using a 37-bus distribution network considering real data. Results proved that the method improves the reactive power management by taking advantage of the full capabilities of the DER and by reducing the injection of reactive power by the TSO in the distribution network and, therefore, reducing losses.