Abstract
Technology driven interventions provide us with an increasing amount of fine-grained data about the patient. This data includes regular ecological momentary assessments (EMA) but also response times to EMA questions by a user. When observing this data, we see a huge variation between the patterns exhibited by different patients. Some are more stable while others vary a lot over time. This poses a challenging problem for the domain of artificial intelligence and makes on wondering whether it is possible to predict the future mental state of a patient using the data that is available. In the end, these predictions could potentially contribute to interventions that tailor the feedback to the user on a daily basis, for example by warning a user that a fall-back might be expected during the next days, or by applying a strategy to prevent the fall-back from occurring in the first place. In this work, we focus on short term mood prediction by considering the adherence and usage data as an additional predictor. We apply recurrent neural networks to handle the temporal aspects best and try to explore whether individual, group level, or one single predictive model provides the highest predictive performance (measured using the root mean squared error (RMSE)). We use data collected from patients from five countries who used the ICT4Depression/MoodBuster platform in the context of the EU E-COMPARED project. In total, we used the data from 143 patients (with between 9 and 425 days of EMA data) who were diagnosed with a major depressive disorder according to DSM-IV. Results show that we can make predictions of short term mood change quite accurate (ranging between 0.065 and 0.11). The past EMA mood ratings proved to be the most influential while adherence and usage data did not improve prediction accuracy. In general, group level predictions proved to be the most promising, however differences were not significant. Short term mood prediction remains a difficult task, but from this research we can conclude that sophisticated machine learning algorithms/setups can result in accurate performance. For future work, we want to use more data from the mobile phone to improve predictive performance of short term mood.

Abstract
Creating an augmented reality (AR) urban tourism application presents several interactivity challenges on how to convey an engaging multimedia experience on-site. This article describes a methodology for fast prototyping of multimedia mobile applications dedicated to urban tourism storytelling with special focus on AR techniques. Following the lessons learned in previous applications the systematic creation of location-based augmented reality (LBAR) applications is explored in this article. The goal is to create serious games for tourism that follow a main narrative but where the story can automatically adapt itself to the current location of the player, assimilate possible detours and allow posterior out-of-location playback. Adaptable stories can use dynamic information from map sources such as points of interest (POI), elevation or virtual buildings. The article discusses and presents solutions for media acquisition, interactive storytelling, game-design interface and multi-disciplinary coordination for mobile app development.

Abstract
We study a mathematical model of immune response by T cells where the regulatory T cells (Treg) inhibit interleukin 2 (IL-2) secretion. We model the suppression of the autoimmune line of T cells after a different line of T cells responded to a pathogen infection. In this paper, we show that if the population of the pathogen responding line of T cells becomes large enough, the competition for IL-2 and the increase in the death rates may lead to a depletion in the concentration of autoimmune T cells. Provided this lasts for a sufficiently long time, the concentration of autoimmune T cells can be brought down to values inside the basin of attraction of the controlled state, and autoimmunity can be suppressed.

Abstract
Directive-drive programming models, such as OpenMP, are one solution for exploiting the potential of multi-core architectures, and enable developers to accelerate software applications by adding annotations on for-type loops and other code regions. However, manual parallelization of applications is known to be a non trivial and time consuming process, requiring parallel programming skills. Automatic parallelization approaches can reduce the burden on the application development side. This paper presents an OpenMP based automatic parallelization compiler, named AutoPar-Clava, for automatic identification and annotation of loops in C code. By using static analysis, parallelizable regions are detected, and a compilable OpenMP parallel code from the sequential version is produced. In order to reduce the accesses to shared memory by each thread, each variable is categorized into the proper OpenMP scoping. Also, AutoPar-Clava is able to support reduction on arrays, which is available since OpenMP 4.5. The effectiveness of AutoPar-Clava is evaluated by means of the Polyhedral Benchmark suite, and targeting a N-cores x86-based computing platform. The achieved results are very promising and compare favorably with closely related auto-parallelization compilers such as Intel C/C++ Compiler (i.e., icc), ROSE, TRACO, and Cetus.

Abstract
CFD simulations are a fundamental engineering application, implying huge workloads, often with dynamic behaviour due to runtime mesh refinement. Parallel processing over heterogeneous distributed memory clusters is often used to process such workloads. The execution of dynamic workloads over a set of heterogeneous resources leads to load imbalances that severely impacts execution time, when static uniform load distribution is used. This paper proposes applying dynamic, heterogeneity aware, load balancing techniques within CFD simulations. nSharma, a software package that fully integrates with OpenFOAM, is presented and assessed. Performance gains are demonstrated, achieved by reducing busy times standard deviation among resources, i.e., heterogeneous computing resources are kept busy with useful work due to an effective workload distribution. To best of authors’ knowledge, nSharma is the first implementation and integration of heterogeneity aware load balancing in OpenFOAM and will be made publicly available in order to foster its adoption by the large community of OpenFOAM users.

Abstract
Multi-label classification is a methodology that tries to solve classification problems where multiple classes are associated with each data example. Data streams pose new challenges to this methodology caused by the massive amounts of structured data production. In fact, most of the existent batch mode methods may not support this condition. Therefore, this paper proposes two multi-label classification methods based on rule and ensembles learning from continuous flow of data. These methods are derived from a multi-target regression algorithm. The main contribution of this work is the rule specialization for subsets of class labels, instead of the usual local (individual models for each output) or a global (one model for all outputs) methods. Prequential evaluation was conducted where global, local and subset operation modes were compared against other online classifiers found in the literature. Six real-world data sets were used. The evaluation demonstrated that the subset specialization presents competitive performance, when compared to local and global approaches and online classifiers found in the literature.