Abstract
Obstructive Sleep Apnea (OSA) affects 2-4% of the population worldwide. The standard test for OSA diagnosis is polysomnography (PSG), an expensive exam limited to urban areas. Furthermore, nearly half of all PSG tests results are negative for OSA. This work aims to reduce these unnecessary exams, by defining an auxiliary diagnostic method that could be used to assess patient's need for PSG, according to their probability of OSA diagnosis. A prospective study was conducted on adult patients with OSA suspicion who performed PSG at our sleep laboratory in Portugal. The studied clinical variables were defined after literature review and collected during consultation. Two comparable cohorts were studied for derivation (n=86) and validation (n=33) of models. Three classifiers were analyzed - a multiple logistic regression classifier (AUC=80.0%) and two Bayesian networks classifiers - Naive Bayes (AUC=81.3%) and Tree Augmented Naive Bayes (TAN, AUC=81.4%) - aiming at the best possible specificity (identification of unnecessary exams). Overall, sensitivity-adjusted models could detect normal patients, preventing unnecessary PSG, while keeping sensitivity high. Furthermore, the graphical representation of TAN can be explored by the physician during consultation, making it a helpful tool to assess patients' need to perform PSG.

Abstract

Abstract
The usage of Autonomous Underwater Vehicles (AUVs), Remotely Operated Vehicles (ROVs), and sensors in surveillance, maintenance and inspection of underwater facilities is increasing the need for broadband, cost-e ective communications solutions. Current solutions, mainly based on acoustic communications, enable long ranges but provide low bitrates and have high communication delays. Despite its strong attenuation underwater, RF is envisioned as a technology to enable broadband, short-range communications. We present an ns-3 and an experimental evaluation of IEEE 802.11 networks in freshwater at 700 MHz, 2.4 GHz and 5 GHz frequency bands. Evaluation results con rm the accuracy of the new developed ns-3 underwater RF propagation model and show that IEEE 802.11 networks are feasible for broadband, short range underwater communications, with range and throughput exceeding 2 m and 100 Mbit/s, respectively.

Abstract
This project has been motivated by the need to standardize, preserve, and share the data sets of the Laboratory of Vibrations and Structural Monitoring (ViBest, www.fe.up.pt/vibest) of FEUP, produced by several long term projects individually managed. The solution presented is meant to support the process of Structural Health Monitoring, offering features to catalogue the projects, their goals and components, to store and visualize their acquired and processed data through time, and to preserve the data in a standardized form for all the research unit and extensible to future applications. The result is a digital archive with automatic ingestion of new data files and a Web interface with access control and tools for information management. There is a batch export functionality to deal with large data transfers. It is being used on monitoring data related with different kinds of structural health monitoring applications. The standardization and preservation of all data sets acquired in multiple applications will be certainly a solid basis for further research, either at a local basis or in the context of international joint cooperation.

Abstract
Linear Meld is a concurrent forward-chaining linear logic programming language where logical facts can be asserted and retracted in a structured way. The database of facts is partitioned by the nodes of a graph structure which leads to parallelism if nodes are executed simultaneously. Communication arises whenever nodes send facts to other nodes by fact derivation. We present an overview of the virtual machine that we implemented to run Linear Meld on multicores, including code organization, thread management, rule execution and database organization for efficient fact insertion, lookup and deletion. Although our virtual machine is a work-in-progress, our results already show that Linear Meld is not only capable of scaling graph and machine learning programs but it also exhibits some interesting performance results when compared against other programming languages.

Abstract
This paper presents an approach that enables the staging of choreographies for education and training purposes in multiple virtual world platforms. Choreography is the description of a set of actions that must or may be executed by a group of participants, including the goals to be achieved and any restrictions that may exist. For capturing and representing multi-actor multidomain choreographies an approach based on ontologies with distinct levels of abstraction is adopted. Further, this paper proposes a modelling driven approach and a set of processes that, through mappings between ontologies, enable the automatic construction of a platform-specific choreography from a platform-independent one, thus reducing the time and effort of the choreography development. For this, the MDA paradigm was adopted and adapted in a way where models can reflect two dimensions of independence: platform independence and application domain independence. We also point the guidelines for staging the choreography in a virtual world platform.