Heterogeneous computing with an algorithmic skeleton framework

The Graphics Processing Unit (GPU) is present in almost every modern day personal computer. Despite its specific purpose design, they have been increasingly used for general computations with very good results. Hence, there is a growing effort from the community to seamlessly integrate this kind of devices in everyday computing. However, to fully exploit the potential of a system comprising GPUs and CPUs, these devices should be presented to the programmer as a single platform. The efficient combination of the power of CPU and GPU devices is highly dependent on each device’s characteristics, resulting in platform specific applications that cannot be ported to different systems. Also, the most efficient work balance among devices is highly dependable on the computations to be performed and respective data sizes. In this work, we propose a solution for heterogeneous environments based on the abstraction level provided by algorithmic skeletons. Our goal is to take full advantage of the power of all CPU and GPU devices present in a system, without the need for different kernel implementations nor explicit work-distribution.To that end, we extended Marrow, an algorithmic skeleton framework for multi-GPUs, to support CPU computations and efficiently balance the work-load between devices. Our approach is based on an offline training execution that identifies the ideal work balance and platform configurations for a given application and input data size. The evaluation of this work shows that the combination of CPU and GPU devices can significantly boost the performance of our benchmarks in the tested environments, when compared to GPU-only executions.

Optimization of breast tomosynthesis image reconstruction using parallel computing

Breast cancer is the most common cancer among women, being a major public health problem. Worldwide, X-ray mammography is the current gold-standard for medical imaging of breast cancer. However, it has associated some well-known limitations. The false-negative rates, up to 66% in symptomatic women, and the false-positive rates, up to 60%, are a continued source of concern and debate. These drawbacks prompt the development of other imaging techniques for breast cancer detection, in which Digital Breast Tomosynthesis (DBT) is included. DBT is a 3D radiographic technique that reduces the obscuring effect of tissue overlap and appears to address both issues of false-negative and false-positive rates. The 3D images in DBT are only achieved through image reconstruction methods. These methods play an important role in a clinical setting since there is a need to implement a reconstruction process that is both accurate and fast. This dissertation deals with the optimization of iterative algorithms, with parallel computing through an implementation on Graphics Processing Units (GPUs) to make the 3D reconstruction faster using Compute Unified Device Architecture (CUDA). Iterative algorithms have shown to produce the highest quality DBT images, but since they are computationally intensive, their clinical use is currently rejected. These algorithms have the potential to reduce patient dose in DBT scans. A method of integrating CUDA in Interactive Data Language (IDL) is proposed in order to accelerate the DBT image reconstructions. This method has never been attempted before for DBT. In this work the system matrix calculation, the most computationally expensive part of iterative algorithms, is accelerated. A speedup of 1.6 is achieved proving the fact that GPUs can accelerate the IDL implementation.

Knowledge management framework based on brain models and human physiology

The life of humans and most living beings depend on sensation and perception for the best assessment of the surrounding world. Sensorial organs acquire a variety of stimuli that are interpreted and integrated in our brain for immediate use or stored in memory for later recall. Among the reasoning aspects, a person has to decide what to do with available information. Emotions are classifiers of collected information, assigning a personal meaning to objects, events and individuals, making part of our own identity. Emotions play a decisive role in cognitive processes as reasoning, decision and memory by assigning relevance to collected information. The access to pervasive computing devices, empowered by the ability to sense and perceive the world, provides new forms of acquiring and integrating information. But prior to data assessment on its usefulness, systems must capture and ensure that data is properly managed for diverse possible goals. Portable and wearable devices are now able to gather and store information, from the environment and from our body, using cloud based services and Internet connections. Systems limitations in handling sensorial data, compared with our sensorial capabilities constitute an identified problem. Another problem is the lack of interoperability between humans and devices, as they do not properly understand human’s emotional states and human needs. Addressing those problems is a motivation for the present research work. The mission hereby assumed is to include sensorial and physiological data into a Framework that will be able to manage collected data towards human cognitive functions, supported by a new data model. By learning from selected human functional and behavioural models and reasoning over collected data, the Framework aims at providing evaluation on a person’s emotional state, for empowering human centric applications, along with the capability of storing episodic information on a person’s life with physiologic indicators on emotional states to be used by new generation applications.

From "Humanities and Computing" to "Digital Humanities": Digital Humanities in Portugal with a focus on Historical Research

In the following text I will develop three major aspects. The first is to draw attention to those who seem to have been the disciplinary fields where, despite everything, the Digital Humanities (in the broad perspective as will be regarded here) have asserted themselves in a more comprehensive manner. I think it is here that I run into greater risks, not only for what I have mentioned above, but certainly because a significant part, perhaps, of the achievements and of the researchers might have escaped the look that I sought to cast upon the past few decades, always influenced by my own experience and the work carried out in the field of History. But this can be considered as a work in progress and it is open to criticism and suggestions. A second point to note is that emphasis will be given to the main lines of development in the relationship between historical research and digital methodologies, resources and tools. Finally, I will try to make a brief analysis of what has been the Digital Humanities discourse appropriation in recent years, with very debatable data and methods for sure, because studies are still scarce and little systematic information is available that would allow to go beyond an introductory reflection.

Evaluating ubiquitous computing environments using 3D simulation

Human activity is very dynamic and subtle, and most physical environments are also highly dynamic and support a vast range of social practices that do not map directly into any immediate ubiquitous computing functionally. Identifying what is valuable to people is very hard and obviously leads to great uncertainty regarding the type of support needed and the type of resources needed to create such support. We have addressed the issues of system development through the adoption of a Crowdsourced software development model [13]. We have designed and developed Anywhere places, an open and flexible system support infrastructure for Ubiquitous Computing that is based on a balanced combination between global services and applications and situated devices. Evaluation, however, is still an open problem. The characteristics of ubiquitous computing environments make their evaluation very complex: there are no globally accepted metrics and it is very difficult to evaluate large-scale and long-term environments in real contexts. In this paper, we describe a first proposal of an hybrid 3D simulated prototype of Anywhere places that combines simulated and real components to generate a mixed reality which can be used to assess the envisaged ubiquitous computing environments [17].

A soft computing approach to kidney diseases evaluation

Kidney renal failure means that one’s kidney have unexpectedly stopped functioning, i.e., once chronic disease is exposed, the presence or degree of kidney dysfunction and its progression must be assessed, and the underlying syndrome has to be diagnosed. Although the patient’s history and physical examination may denote good practice, some key information has to be obtained from valuation of the glomerular filtration rate, and the analysis of serum biomarkers. Indeed, chronic kidney sickness depicts anomalous kidney function and/or its makeup, i.e., there is evidence that treatment may avoid or delay its progression, either by reducing and prevent the development of some associated complications, namely hypertension, obesity, diabetes mellitus, and cardiovascular complications. Acute kidney injury appears abruptly, with a rapid deterioration of the renal function, but is often reversible if it is recognized early and treated promptly. In both situations, i.e., acute kidney injury and chronic kidney disease, an early intervention can significantly improve the prognosis.The assessment of these pathologies is therefore mandatory, although it is hard to do it with traditional methodologies and existing tools for problem solving. Hence, in this work, we will focus on the development of a hybrid decision support system, in terms of its knowledge representation and reasoning procedures based on Logic Programming, that will allow one to consider incomplete, unknown, and even contradictory information, complemented with an approach to computing centered on Artificial Neural Networks, in order to weigh the Degree-of-Confidence that one has on such a happening. The present study involved 558 patients with an age average of 51.7 years and the chronic kidney disease was observed in 175 cases. The dataset comprise twenty four variables, grouped into five main categories. The proposed model showed a good performance in the diagnosis of chronic kidney disease, since the sensitivity and the specificity exhibited values range between 93.1 and 94.9 and 91.9–94.2 %, respectively.

Immersiveness of ubiquitous computing environments prototypes: A case study

The development of ubiquitous computing (ubicomp) environments raises several challenges in terms of their evaluation. Ubicomp virtual reality prototyping tools enable users to experience the system to be developed and are of great help to face those challenges, as they support developers in assessing the consequences of a design decision in the early phases of development. Given the situated nature of ubicomp environments, a particular issue to consider is the level of realism provided by the prototypes. This work presents a case study where two ubicomp prototypes, featuring different levels of immersion (desktop-based versus CAVE-based), were developed and compared. The goal was to determine the cost/benefits relation of both solutions, which provided better user experience results, and whether or not simpler solutions provide the same user experience results as more elaborate one.

Filososofía y Ciencia Computacional: Aspectos epistemológicos, ontológicos y metodológicos. Philosophy and Computing Sciencs: Epistemological, Ontological and Methodological Aspects.

Uno de los temas centrales del proyecto concierne la naturaleza de la ciencia de la computación. La reciente aparición de esta disciplina sumada a su origen híbrido como ciencia formal y disciplina tecnológica hace que su caracterización aún no esté completa y menos aún acordada entre los científicos del área. En el trabajo Three paradigms of Computer Science de A. Eden, se presentan tres posiciones admitidamente exageradas acerca de como entender tanto el objeto de estudio (ontología) como los métodos de trabajo (metodología) y la estructura de la teoría y las justificaciones del conocimiento informático (epistemología): La llamada racionalista, la cual se basa en la idea de que los programas son fórmulas lógicas y que la forma de trabajo es deductiva, la tecnocrática que presenta a la ciencia computacional como una disciplina ingenieril y la ahi llamada científica, la cual asimilaría a la computación a las ciencias empíricas. Algunos de los problemas de ciencia de la computación están relacionados con cuestiones de filosofía de la matemática, en particular la relación entre las entidades abstractas y el mundo. Sin embargo, el carácter prescriptivo de los axiomas y teoremas de las teorías de la programación puede permitir interpretaciones alternativas y cuestionaría fuertemente la posibilidad de pensar a la ciencia de la computación como una ciencia empírica, al menos en el sentido tradicional. Por otro lado, es posible que el tipo de análisis aplicado a las ciencias de la computación propuesto en este proyecto aporte nuevas ideas para pensar problemas de filosofía de la matemática. Un ejemplo de estos posibles aportes puede verse en el trabajo de Arkoudas Computers, Justi?cation, and Mathematical Knowledge el cual echa nueva luz al problema del significado de las demostraciones matemáticas.Los objetivos del proyecto son: Caracterizar el campo de las ciencias de la computación.Evaluar los fundamentos ontológicos, epistemológicos y metodológicos de la ciencia de la computación actual.Analizar las relaciones entre las diferentes perspectivas heurísticas y epistémicas y las practicas de la programación.

monitoring-based approach to object-oriented real-time computing

Monitoring, object-orientation, real-time, execution-time, scheduling

The 2004 election in Spain: terrorism, accountability, and voting

In this paper the electoral consequences of the Islamist terrorist attacks on March 11, 2004 are analysed. According to a quantitative analysis based on a post-electoral survey, we show the causal mechanisms that transform voters’ reactions to the bombings into a particular electoral behaviour and estimate their relevance in the electoral results on March 14, 2004

Plataforma Computing@home

En este proyecto se han visto dos sistemas de computación distribuida diferentes entre ellos: Condor y BOINC. Se exploran las posibilidades para poder conseguir que ambos sistemas logren trabajar conjuntamente, escogiendo la parte más efectiva de cada uno de los sistemas con el fin de complementarse.

Computing Markov-Perfect Optimal Policies in Business-Cycle Models

Time-inconsistency is an essential feature of many policy problems (Kydland and Prescott, 1977). This paper presents and compares three methods for computing Markov-perfect optimal policies in stochastic nonlinear business cycle models. The methods considered include value function iteration, generalized Euler-equations, and parameterized shadow prices. In the context of a business cycle model in which a scal authority chooses government spending and income taxation optimally, while lacking the ability to commit, we show that the solutions obtained using value function iteration and generalized Euler equations are somewhat more accurate than that obtained using parameterized shadow prices. Among these three methods, we show that value function iteration can be applied easily, even to environments that include a risk-sensitive scal authority and/or inequality constraints on government spending. We show that the risk-sensitive scal authority lowers government spending and income-taxation, reducing the disincentive households face to accumulate wealth.

Multi-level SLA-driven resource management on Cloud Providers

Nowadays, service providers in the Cloud offer complex services ready to be used as it was a commodity like water or electricity to their customers with any other extra effort for them. However, providing these services implies a high management effort which requires a lot of human interaction. Furthermore, an efficient resource management mechanism considering only provider's resources is, though necessary, not enough, because the provider's profit is limited by the amount of resources it owns. Dynamically outsourcing resources to other providers in response to demand variation avoids this problem and makes the provider to get more profit. A key technology for achieving these goals is virtualization which facilitates provider's management and provides on-demand virtual environments, which are isolated and consolidated in order to achieve a better utilization of the provider's resources. Nevertheless, dealing with some virtualization capabilities implies an effort for the user in order to take benefit from them. In order to avoid this problem, we are contributing the research community with a virtualized environment manager which aims to provide virtual machines that fulfils with the user requirements. Another challenge is sharing resources among different federated Cloud providers while exploiting the features of virtualization in a new approach for facilitating providers' management. This project aims for reducing provider's costs and at the same time fulfilling the quality of service agreed with the customers while maximizing the provider's revenue. It considers resource management at several layers, namely locally to each node in the provider, among different nodes in the provider, and among different federated providers. This latter layer supports the novel capabilities of outsourcing when the local resources are not enough to fulfil the users demand, and offering resources to other providers when the local resources are underused.