000 Informatik, Informationswissenschaft, allgemeine Werke
Refine
Year of publication
Document Type
- Article (245)
- Part of Periodical (37)
- Doctoral Thesis (22)
- Book (19)
- Review (9)
- Conference Proceeding (7)
- Periodical (4)
- Part of a Book (2)
- Other (2)
- Contribution to a Periodical (1)
Is part of the Bibliography
- no (349)
Keywords
- Frankfurt <Main> / Universität (115)
- Frankfurt <Main> (60)
- Frankfurt (42)
- Forschung (30)
- Zeitschrift (29)
- Biographie (26)
- Preis <Auszeichnung> (10)
- Wissenschaft (10)
- Frankfurt / Universität (9)
- Adorno (7)
Institute
- Präsidium (42)
- Informatik (28)
- Fachübergreifend (16)
- Frankfurt Institute for Advanced Studies (FIAS) (10)
- Informatik und Mathematik (3)
- Biowissenschaften (2)
- Erziehungswissenschaften (2)
- Evangelische Theologie (2)
- Extern (2)
- Geschichtswissenschaften (2)
Ausgehend von der grundlegenden narratologischen Bedeutung von Ereignissen in ihrer konstitutiven Funktion für Erzähltexte wurde im Projekt "Evaluating Events in Narrative Theory" (EvENT) ein Ansatz entwickelt, mit welchem Ereignisse auf der Textoberfläche und daher maschinenlesbar modelliert werden können. Dieser Beitrag skizziert bisherige Arbeitsschritte und ausgewählte Ergebnisse des Projekts, zu denen die Generierung von Narrativitätsgraphen zählt.
Only a few studies on the nocturnal behavior of African ungulates exist so far, with mostly small sample sizes. For a comprehensive understanding of nocturnal behavior, the data basis needs to be expanded. Results obtained by observing zoo animals can provide clues for the study of wild animals and furthermore contribute to a better understanding of animal welfare and better husbandry conditions in zoos. The current contribution reduces the lack of data in two ways. First, we present a stand-alone open-source software package based on deep learning techniques, named Behavioral Observations by Videos and Images using Deep-Learning Software (BOVIDS). It can be used to identify ungulates in their enclosure and to determine the three behavioral poses “Standing,” “Lying—head up,” and “Lying—head down” on 11,411 h of video material with an accuracy of 99.4%. Second, BOVIDS is used to conduct a case study on 25 common elands (Tragelaphus oryx) out of 5 EAZA zoos with a total of 822 nights, yielding the first detailed description of the nightly behavior of common elands. Our results indicate that age and sex are influencing factors on the nocturnal activity budget, the length of behavioral phases as well as the number of phases per behavioral state during the night while the keeping zoo has no significant influence. It is found that males spend more time in REM sleep posture than females while young animals spend more time in this position than adult ones. Finally, the results suggest a rhythm between the Standing and Lying phases among common elands that opens future research directions.
Artificial Intelligence (AI) and Machine Learning (ML) are currently hot topics in industry and business practice, while management-oriented research disciplines seem reluctant to adopt these sophisticated data analytics methods as research instruments. Even the Information Systems (IS) discipline with its close connections to Computer Science seems to be conservative when conducting empirical research endeavors. To assess the magnitude of the problem and to understand its causes, we conducted a bibliographic review on publications in high-level IS journals. We reviewed 1,838 articles that matched corresponding keyword-queries in journals from the AIS senior scholar basket, Electronic Markets and Decision Support Systems (Ranked B). In addition, we conducted a survey among IS researchers (N = 110). Based on the findings from our sample we evaluate different potential causes that could explain why ML methods are rather underrepresented in top-tier journals and discuss how the IS discipline could successfully incorporate ML methods in research undertakings.
Eine Erkrankung zählt in der Europäischen Union zu den Seltenen Erkrankungen (SE), wenn diese nicht mehr als 5 von 10.000 Menschen betrifft. Derzeit existiert mit mehr als 6000 SE eine sowohl große als auch heterogene Menge an unterschiedlichen Krankheitsbilder, die in ihrer Symptomatik komplex, vielschichtig und damit im medizinischen Alltag schwierig einzuordnen sind. Dies erschwert Diagnosefindung und Behandlung sowie das Auffinden eines passenden Ansprechpartners, da es nur wenige Experten für jede einzelne SE gibt. Der medizinische Versorgungsatlas für Seltene Erkrankungen www.se-atlas.de ermöglicht anhand von Erkrankungsnamen die Suche nach Versorgungseinrichtungen und Selbsthilfeorganisationen zu bestimmten SE und stellt die Suchergebnisse geografisch dar. Ebenso gibt er einen Überblick über alle deutschen Zentren für SE, die eine Anlaufstelle für betroffene Personen mit unklarer Diagnose darstellen. Der se-atlas dient als Kompass durch die heterogene Menge an Informationen über Versorgungseinrichtungen für SE und stellt niederschwellig Informationen für eine breite Nutzergruppe von Betroffenen bis hin zu Mitgliedern des medizinischen Versorgungsteams bereit.
OPEN SCIENCE, VERSION 3.0: Breaking down barriers for equitable and efficient research communication
(2022)
Wissenschaft
(2022)
Georg Toepfer wendet sich Ganzheitsidealen der Wissenschaft und ihrer Theorie von der Antike bis zur zeitgenössischen Wissensgeschichte zu. Ganzheit kann dabei auf unterschiedlichen Ebenen zum Tragen kommen. Die Wissenschaft beschäftigt sich mit ihrer inneren Einheit auf methodologischer Ebene, sie fragt nach der Ganzheit oder Ganzheitlichkeit ihrer Gegenstände (im Rahmen etwa einer Überwindung des Leib-Seele-Dualismus), sie kann eine Summe aller Wissenschaften als Ganzheit in Aussicht stellen oder den Versuch unternehmen, eine Einheit oder Ganzheit der gesamten Menschheit zu befördern. Insgesamt sei festzuhalten, dass Forderungen nach einer Einheit der Wissenschaft verstärkt zu Zeiten aufkämen, in denen sich epistemologisch wie praktisch das exakte Gegenteil beobachten lasse. Auch wirke die Ganzheitsrhetorik der Wissenschaft "mitnichten integrativ", sondern arbeite im Gegen teil oft "mit massiven Ausgrenzungen". Das Ganzheitsproblem habe die wissenschaftliche Selbstreflexion oft auch hinsichtlich ihrer eigenen Visualisierbarkeit beschäftigt. Dies findet seinen Niederschlag in frühneuzeitlichen Diagrammtypen wie dem Baum, dem Haus oder dem Bergwerk, in der zweidimensionalen Landkarte der 'Encyclopédie' oder in einer von Jean Piaget noch im 20. Jahrhundert bearbeiteten Kreisfigur. Solche Formen des Ganzen offenbarten die "Simplifikation" des wissenschaftstheoretischen Zugriffs und müssten oft ganze Wissensbereiche ausschließen; im Fall Piagets etwa die kompletten Geisteswissenschaften.
Die drei Gewinner des ersten Digital Publishing Creative Ideas-Wettbewerb wurden auf der Frankfurter Buchmesse 2013 bekanntgegeben. Der Wettbewerb wurde erstmals in einer Kooperation zwischen dem Goethe-Unibator (Existenzgründerzentrum der Goethe-Universität) und der Frankfurt Academy (als Teil der Buchmesse) ins Leben gerufen. Neben dem „Book of the Future“ von Gottfried Vossen und Florian Stahl konnten folgende Projekte reüssieren: Andreas Dollmayer (Stuttgart) setzte sich mit der Idee für einen interaktiven Reiseführer durch. Auf der Basis digitaler Karten- und Bilderdienste sowie Blogs und in Verbindung mit Social Media entstehen intelligente Reiseführer in verschiedensten Ausgabeformaten, an die sich Reiseanbieter andocken können. Hiroki Kamata (Tokio) wiederum behauptete sich mit einer Projektidee, bei der antike Bücher, die vor der Zeit Gutenbergs hergestellt wurden, auf der Basis einer digitalen Arbeitsumgebung wieder zum Leben erweckt werden.
[Es] zeichnen sich drei grundlegend unterscheidbare Richtungen ab, die als Aufgabenfeld einer Computerphilologie gesehen werden: die Entwicklung digitaler Editionen, die Erforschung PC-basierter Literatur und die Ausbildung PC-basierter Methoden einschließlich der Entwicklung und Programmierung entsprechender Anwendungen. Diese Definition bietet Anlass zur Reflexion über die Frage nach Methodologie und Disziplinarität der Computerphilologie. Die Aufzählung […] kann jedoch nur den Ausgangspunkt markieren, da sie lediglich den Einsatzbereich des PCs absteckt. Aber darauf aufbauend lassen sich drei Thesen zur näheren Bestimmung computerphilologischer Arbeit formulieren, die nachfolgend zur Diskussion gestellt werden.
Türkçe, Almanca, İngilizce gibi doğal dillerde bir tümce temelde özne ve yüklemden oluşur. Benzer şekilde biçimsel dillerde de bir tümce, yüklem ve argümandan oluşur. Yüklemler P, Q, R gibi büyük harflerle, argümanlar ise x, y, z gibi küçük harflerle gösterilir. Örneğin olumlu bir tümce P(x), olumsuz bir tümce ise -P(x) şeklinde ifade edilebilir. Ancak bazen bir tümcenin olumlu mu yoksa olumsuz mu olduğu net bir şekilde belli olmayabilir. Bu tür durumlarda mevcut sembolik gösterimde belirsizlikler ortaya çıkabilmektedir. Olumlu tümcelere matematiksel olarak 1, olumsuz tümcelere ise 0 değerinin verildiği varsayılırsa, olumluluk veya olumsuzluk durumu belirsiz olan tümceler ancak bu iki değer arasında bir değer alabilir. Diğer bir deyişle P(x) şeklinde gösterilebilen bir tümceyi P1(x), -P(x) şeklinde gösterilen bir tümceyi ise P0(x) şeklinde ifade etmek mümkündür; fakat olumluluğu kesin olmayan tümceler bu değerlerle gösterilemeyeceği için başka bir ifade şekline ihtiyaç vardır. Çünkü bu tümcelerdeki iş, oluş veya hareketin gerçekleşme oranı ne 0 ne de 1'dir; 0 ve 1 arasında bir değerdir. Bu çalışmada bu tür tümcelerin biçimsel dillerde nasıl ifade edilebileceğine dair bir öneride bulunmak ve bulanık küme kuramıyla olumluluğu derecelendirmek amaçlanmıştır. Bu amaç doğrultusunda önerilen yaklaşım birtakım örnek tümceler üzerinde uygulanmış ve söz konusu tümceler bulanık sembolik bir gösterimle ifade edilmiştir.
Why do we need to communicate science? Is science, with its highly specialised language and its arcane methods, too distant to be understood by the public? Is it really possible for citizens to participate meaningfully in scientific research projects and debate? Should scientists be mandated to engage with the public to facilitate better understanding of science? How can they best communicate their special knowledge to be intelligible? These and a plethora of related questions are being raised by researchers and politicians alike as they have become convinced that science and society need to draw nearer to one another. Once the persuasion took hold that science should open up to the public and these questions were raised, it became clear that coming up with satisfactory answers would be a complex challenge. The inaccessibility of scientific language and methods, due to ever increasing specialisation, is at the base of its very success. Thus, translating specialised knowledge to become understandable, interesting and relevant to various publics creates particular perils. This is exacerbated by the ongoing disruption of the public discourse through the digitisation of communication platforms. For example, the availability of medical knowledge on the internet and the immense opportunities to inform oneself about health risks via social media are undermined by the manipulable nature of this technology that does not allow its users to distinguish between credible content and misinformation. In countries around the world, scientists, policy-makers and the public have high hopes for science communication: that it may elevate its populations educationally, that it may raise the level of sound decision-making for people in their daily lives, and that it may contribute to innovation and economic well-being. This collection of current reflections gives an insight into the issues that have to be addressed by research to reach these noble goals, for South Africa and by South Africans in particular.
Modern-day science is under great pressure. A potent mix of increasing expectations, limited resources, tensions between competition and cooperation, and the need for evidence-based funding is creating major change in how science is conducted and perceived. Amidst this 'perfect storm' is the allure of 'research excellence', a concept that drives decisions made by universities and funders, and defines scientists' research strategies and career trajectories. But what is 'excellent' science? And how to recognise it? After decades of inquiry and debate there is still no satisfactory answer. Are we asking the wrong question? Is reality more complex, and 'excellence in science' more elusive, than many are willing to admit? And how should excellence be defined in different parts of the world, particularly in lower-income countries of the 'Global South' where science is expected to contribute to pressing development issues, despite often scarce resources? Many wonder whether the Global South is importing, with or without consenting, the flawed tools for research evaluation from North America and Europe that are not fit for purpose. This book takes a critical view of these issues, touching on conceptual issues and practical problems that inevitably emerge when 'excellence' is at the center of science systems. Emerging from the capacity-building work of the Science Granting Councils Initiative in sub-Saharan Africa, it speaks to scholars, as well as to managers and funders of research around the world. Confronting sticky problems and uncomfortable truths, the chapters contain insights and recommendations that point towards new solutions - both for the Global South and the Global North.
Biological ageing is a degenerative and irreversible process, ultimately leading to death of the organism. The process is complex and under the control of genetic, environmental and stochastic traits. Although many theories have been established during the last decades, none of these are able to fully describe the complex mechanisms, which lead to ageing. Generally, biological processes and environmental factors lead to molecular damage and an accumulation of impaired cellular components. In contrast, counteracting surveillance systems are effective, including repair, remodelling and degradation of damaged or impaired components, respectively. Nevertheless, at some point these systems are no longer effective, either because the increasing amount of molecular damages can not longer be removed efficiently or because the repairing and removing mechanisms themselves become affected by impairing effects. The organism finally declines and dies. To investigate and to understand these counteracting mechanisms and the complex interplay of decline and maintenance, holistic and systems biological investigations are required. Hence, the processes which lead to ageing in the fungal model organism Podospora anserina, had been analysed using different advanced bioinformatics methods. In contrast to many other ageing models, P. anserina exhibits a short lifespan, a less biochemical complexity and it provides a good accessibility for genetic manipulations.
To achieve a general overview on the different biochemical processes, which are affected during ageing in P. anserina, an initial comprehensive investigation was applied, which aimed to reveal genes significantly regulated and expressed in an age-dependent manner. This investigation was based on an age-dependent transcriptome analysis. Sophisticated and comprehensive analyses revealed different age-related pathways and indicated that especially autophagy may play a crucial role during ageing. For example, it was found that the expression of autophagy-associated genes increases in the course of ageing.
Subsequently, to investigate and to characterise the autophagy pathway, its associated single components and their interactions, Path2PPI, a new bioinformatics approach, was developed. Path2PPI enables the prediction of protein-protein interaction networks of particular pathways by means of a homology comparison approach and was applied to construct the protein-protein interaction network of autophagy in P. anserina.
The predicted network was extended by experimental data, comprising the transcriptome data as well as newly generated protein-protein interaction data achieved from a yeast two-hybrid analysis. Using different mathematical and statistical methods the topological properties of the constructed network had been compared with those of randomly generated networks to approve its biological significance. In addition, based on this topological and functional analysis, the most important proteins were determined and functional modules were identified, which correspond to the different sub-pathways of autophagy. Due to the integrated transcriptome data the autophagy network could be linked to the ageing process. For example, different proteins had been identified, which genes are continuously up- or down-regulated during ageing and it was shown for the first time that autophagy-associated genes are significantly often co-expressed during ageing.
The presented biological network provides a systems biological view on autophagy and enables further studies, which aim to analyse the relationship of autophagy and ageing. Furthermore, it allows the investigation of potential methods for intervention into the ageing process and to extend the healthy lifespan of P. anserina as well as of other eukaryotic organisms, in particular humans.
We study Gaifman locality and Hanf locality of an extension of first-order logic with modulo p counting quantifiers (FO+MODp , for short) with arbitrary numerical predicates. We require that the validity of formulas is independent of the particular interpretation of the numerical predicates and refer to such formulas as arb-invariant formulas. This paper gives a detailed picture of locality and non-locality properties of arb-invariant FO+MODp . For example, on the class of all finite structures, for any p 2, arb-invariant FO+MODp is neither Hanf nor Gaifman local with respect to a sublinear locality radius. However, in case that p is an odd prime power, it is weakly Gaifman local with a polylogarithmic locality radius. And when restricting attention to the class of string structures, for odd prime powers p, arb-invariant FO+MODp is both Hanf and Gaifman local with a polylogarithmic locality radius. Our negative results build on examples of order-invariant FO+MODp formulas presented in Niemist ̈o’s PhD thesis. Our positive results make use of the close connection between FO+MODp and Boolean circuits built from NOT-gates and AND-, OR-, and MOD p - gates of arbitrary fan-in.
We consider the isolated spelling error correction problem as a specific subproblem of the more general string-to-string translation problem. In this context, we investigate four general string-to-string transformation models that have been suggested in recent years and apply them within the spelling error correction paradigm. In particular, we investigate how a simple ‘k-best decoding plus dictionary lookup’ strategy performs in this context and find that such an approach can significantly outdo baselines such as edit distance, weighted edit distance, and the noisy channel Brill and Moore model to spelling error correction. We also consider elementary combination techniques for our models such as language model weighted majority voting and center string combination. Finally, we consider real-world OCR post-correction for a dataset sampled from medieval Latin texts.
This is a short summary of a recent survey [FR03] focusing on the observed evidence, that Internet connectivity is positively correlated with spread of democracy at high levels of significance. The results of multivariate correlation analysis and probabilities regression estimate models are based on the combined analysis of mid - 1991’s, to 2001 data series of the Eurostat’s and US Census Bureau, the World Bank, and OECD’s statistical data service which track the growth of information technology and rating of freedom and democracy worldwide.
We present an implementation of an interpreter LRPi for the call-by-need calculus LRP, based on a variant of Sestoft's abstract machine Mark 1, extended with an eager garbage collector. It is used as a tool for exact space usage analyses as a support for our investigations into space improvements of call-by-need calculi.
Magnetoencephalography (MEG) measures neural activity non-invasively and at an excellent temporal resolution. Since its invention (Cohen, 1968, 1972), MEG has proven a most valuable tool in neurocognitive (Salmelin et al., 1994) and clinical research (Stufflebeam et al., 2009; Van ’t Ent et al., 2003). MEG is able to measure rapid changes in electrophysiological neural signals related to sensory and cognitive processes. The magnetic fields measured outside the head by MEG directly reflect the cortical currents generated by the synchronised activity of thousands of neuronal sources. This distinguishes MEG from functional magnetic resonance imaging (fMRI), where measurements are only indirectly related to electrophysiological activity through neurovascular coupling...
Already today modern driver assistance systems contribute more and more to make individual mobility in road traffic safer and more comfortable. For this purpose, modern vehicles are equipped with a multitude of sensors and actuators which perceive, interpret and react to the environment of the vehicle. In order to reach the next set of goals along this path, for example to be able to assist the driver in increasingly complex situations or to reach a higher degree of autonomy of driver assistance systems, a detailed understanding of the vehicle environment and especially of other moving traffic participants is necessary.
It is known that motion information plays a key role for human object recognition [Spelke, 1990]. However, full 3D motion information is mostly not taken into account for Stereo Vision-based object segmentation in literature. In this thesis, novel approaches for motion-based object segmentation of stereo image sequences are proposed from which a generic environmental model is derived that contributes to a more precise analysis and understanding of the respective traffic scene. The aim of the environmental model is to yield a minimal scene description in terms of a few moving objects and stationary background such as houses, crash barriers or parking vehicles. A minimal scene description aggregates as much information as possible and it is characterized by its stability, precision and efficiency.
Instead of dense stereo and optical flow information, the proposed object segmentation builds on the so-called Stixel World, an efficient superpixel-like representation of space-time stereo data. As it turns out this step substantially increases stability of the segmentation and it reduces the computational time by several orders of magnitude, thus enabling real-time automotive use in the first place. Besides the efficient, real-time capable optimization, the object segmentation has to be able to cope with significant noise which is due to the measurement principle of the used stereo camera system. For that reason, in order to obtain an optimal solution under the given extreme conditions, the segmentation task is formulated as a Bayesian optimization problem which allows to incorporate regularizing prior knowledge and redundancies into the object segmentation.
Object segmentation as it is discussed here means unsupervised segmentation since typically the number of objects in the scene and their individual object parameters are not known in advance. This information has to be estimated from the input data as well.
For inference, two approaches with their individual pros and cons are proposed, evaluated and compared. The first approach is based on dynamic programming. The key advantage of this approach is the possibility to take into account non-local priors such as shape or object size information which is impossible or which is prohibitively expensive with more local, conventional graph optimization approaches such as graphcut or belief propagation.
In the first instance, the Dynamic Programming approach is limited to one-dimensional data structures, in this case to the first Stixel row. A possible extension to capture multiple Stixel rows is discussed at the end of this thesis.
Further novel contributions include a special outlier concept to handle gross stereo errors associated with so-called stereo tear-off edges. Additionally, object-object interactions are taken into account by explicitly modeling object occlusions. These extensions prove to be dramatic improvements in practice.
This first approach is compared with a second approach that is based on an alternating optimization of the Stixel segmentation and of the relevant object parameters in an expectation maximization (EM) sense. The labeling step is performed by means of the _−expansion graphcut algorithm, the parameter estimation step is done via one-dimensional sampling and multidimensional gradient descent. By using the Stixel World and due to an efficient implementation, one step of the optimization only takes about one millisecond on a standard single CPU core. To the knowledge of the author, at the time of development there was no faster global optimization in a demonstrator car.
For both approaches, various testing scenarios have been carefully selected and allow to examine the proposed methods thoroughly under different real-world conditions with limited groundtruth at hand. As an additional innovative application, the first approach was successfully implemented in a demonstrator car that drove the so-called Bertha Benz Memorial Route from Mannheim to Pforzheim autonomously in real traffic.
At the end of this thesis, the limits of the proposed systems are discussed and a prospect on possible future work is given.
The behaviour of electronic circuits is influenced by ageing effects. Modelling the behaviour of circuits is a standard approach for the design of faster, smaller, more reliable and more robust systems. In this thesis, we propose a formalization of robustness that is derived from a failure model, which is based purely on the behavioural specification of a system. For a given specification, simulation can reveal if a system does not comply with a specification, and thus provide a failure model. Ageing usually works against the specified properties, and ageing models can be incorporated to quantify the impact on specification violations, failures and robustness. We study ageing effects in the context of analogue circuits. Here, models must factor in infinitely many circuit states. Ageing effects have a cause and an impact that require models. On both these ends, the circuit state is highly relevant, an must be factored in. For example, static empirical models for ageing effects are not valid in many cases, because the assumed operating states do not agree with the circuit simulation results. This thesis identifies essential properties of ageing effects and we argue that they need to be taken into account for modelling the interrelation of cause and impact. These properties include frequency dependence, monotonicity, memory and relaxation mechanisms as well as control by arbitrary shaped stress levels. Starting from decay processes, we define a class of ageing models that fits these requirements well while remaining arithmetically accessible by means of a simple structure.
Modeling ageing effects in semiconductor circuits becomes more relevant with higher integration and smaller structure sizes. With respect to miniaturization, digital systems are ahead of analogue systems, and similarly ageing models predominantly focus on digital applications. In the digital domain, the signal levels are either on or off or switching in between. Given an ageing model as a physical effect bound to signal levels, ageing models for components and whole systems can be inferred by means of average operation modes and cycle counts. Functional and faithful ageing effect models for analogue components often require a more fine-grained characterization for physical processes. Here, signal levels can take arbitrary values, to begin with. Such fine-grained, physically inspired ageing models do not scale for larger applications and are hard to simulate in reasonable time. To close the gap between physical processes and system level ageing simulation, we propose a data based modelling strategy, according to which measurement data is turned into ageing models for analogue applications. Ageing data is a set of pairs of stress patterns and the corresponding parameter deviations. Assuming additional properties, such as monotonicity or frequency independence, learning algorithm can find a complete model that is consistent with the data set. These ageing effect models decompose into a controlling stress level, an ageing process, and a parameter that depends on the state of this process. Using this representation, we are able to embed a wide range of ageing effects into behavioural models for circuit components. Based on the developed modelling techniques, we introduce a novel model for the BTI effect, an ageing effect that permits relaxation. In the following, a transistor level ageing model for BTI that targets analogue circuits is proposed. Similarly, we demonstrate how ageing data from analogue transistor level circuit models lift to purely behavioural block models. With this, we are the first to present a data based hierarchical ageing modeling scheme. An ageing simulator for circuits or system level models computes long term transients, solutions of a differential equation. Long term transients are often close to quasi-periodic, in some sense repetitive. If the evaluation of ageing models under quasi-periodic conditions can be done efficiently, long term simulation becomes practical. We describe an adaptive two-time simulation algorithm that basically skips periods during simulation, advancing faster on a second time axis. The bottleneck of two-time simulation is the extrapolation through skipped frames. This involves both the evaluation of the ageing models and the consistency of the boundary conditions. We propose a simulator that computes long term transients exploiting the structure of the proposed ageing models. These models permit extrapolation of the ageing state by means of a locally equivalent stress, a sort of average stress level. This level can be computed efficiently and also gives rise to a dynamic step control mechanism. Ageing simulation has a wide range of applications. This thesis vastly improves the applicability of ageing simulation for analogue circuits in terms of modelling and efficiency. An ageing effect model that is a part of a circuit component model accounts for parametric drift that is directly related to the operation mode. For example asymmetric load on a comparator or power-stage may lead to offset drift, which is not an empiric effect. Monitor circuits can report such effects during operation, when they become significant. Simulating the behaviour of these monitors is important during their development. Ageing effects can be compensated using redundant parts, and annealing can revert broken components to functional. We show that such mechanisms can be simulated in place using our models and algorithms. The aim of automatized circuit synthesis is to create a circuit that implements a specification for a certain use case. Ageing simulation can identify candidates that are more reliable. Efficient ageing simulation allows to factor in various operation modes and helps refining the selection. Using long term ageing simulation, we have analysed the fitness of a set of synthesized operational amplifiers with similar properties concerning various use cases. This procedure enables the selection of the most ageing resilient implementation automatically.
Die in den letzten Jahrzehnten erfolgten Entwicklungen im Bereich der Informations- und Kommunikationstechnologien (IKT) haben beinahe alle Teilbereiche der Gesellschaft erreicht (OECD, 2000) und so können Alltag, Beruf, aber auch soziale Interaktionen benannt werden (Autor, Levy, & Murnane, 2003). Das technologiebasierte Problemlösen stellt ein relevantes Kompetenzkonstrukt dar, wobei die empirische Validierung der Testwertinterpretation noch aussteht. „Technologiebasiertes Problemlösen ist die Kompetenz, digitale Technologien, Kommunikationshilfen und Netzwerke erfolgreich für die Suche, Vermittlung und Interpretation von Informationen zu nutzen.“(OECD, 2009; zitiert nach Rammstedt, 2013). Ziel einer Konstruktvalidierung ist es theoretische Annahmen – sowie nomologische Netze – mit empirischen Belegen zu prüfen und somit Aussagen über die Gültigkeit der Testwertinterpretationen treffen zu können (vgl. Messick, 1995). Zu diesem Zweck wurden die folgenden fünf Leithypothesen aufgestellt:
1. Im technologiebasierten Problemlösen können Teilsequenzen durch Routinen im Umgang mit IKT bewältigt werden.
Mit dem Ziel der Analyse von Lösungsprozessen wurden Prozessdaten aus dem Feldtest der PIAAC-Studie genutzt und automatisierbare Teilschritte (beispielsweise Schließen eines Popups) analysiert. Eine schnelle Bearbeitung dieser Teilschritte wurde auch mit einer höheren Lösungswahrscheinlichkeit des Problems assoziiert. Die Testwerte des technologiebasierten Problemlösens spiegeln die bildungsbiographischen Lernerfolge wider, die zu einem routinierten Umgang mit IKT führen.
2. Durch die Entstehungsgeschichte von IKT begründet, bestehen Differenzen im technologiebasierten Problemlösen.
Entstehung moderner IKT legt eine Differenzierung von drei Kohorten nahe (orientiert an lernintensiven Phasen der Kindheit und Jugend):
- Erlebten die Entstehung und Verbreitung von Hardware (geboren 1946-1966)
- Begleiteten die Verbreitung von Computersoftware und dessen Nutzung (geboren 1966-1981)
- Nutzung und Gestaltung digitaler Inhalte des Internets (geboren 1981-1995)
Erwartungsgemäß unterscheiden sich die Kohorten im technologiebasierten Problemlösen – genauer in der Nutzung von Routinen im Lösungsprozess und im erreichten Kompetenzniveau. Ältere Kohorten benötigten im Mittel mehr Zeit für Teilschritte, die durch Routinen bearbeitet werden können und sie erreichten im Mittel niedrigere Kompetenzniveaus. Obwohl Lesekompetenzen eine wesentliche Bedeutung im technologiebasierten Lösen von Problemen haben, können sie die Kohortenunterschiede nicht umfassend aufklären. Weil ältere Personenkohorten während ihrer formellen Ausbildungsphasen IKT – wie sie heute verwendet werden – nicht nutzen konnten, haben informelle Lerngelegenheiten eine wesentliche Bedeutung. So profitieren ältere Personen von einem regelmäßigen, beruflichen oder privaten Umgang mit diesen Technologien. Die Testwerte des technologiebasierten Problemlösens spiegeln entstehungsgeschichtlich begründete Kohortenunterschiede wider, welche durch lebenslange, formelle und informelle Lerngelegenheiten minimiert werden können.
3. Der Umgang mit IKT ist – begründet durch deren Entstehungsgeschichte – eher weniger Teil der formellen und schulischen Bildung.
Technologiebasiertes Problemlösen wird weniger in formellen und schulischen Lerngelegenheiten erworben, als andere Kompetenzen wie beispielsweise mathematische und Lesekompetenzen. Folglich wurden für das technologiebasierte Problemlösen kleinere Differenzen zwischen hohen, mittleren oder niedrigen Bildungsabschlüssen erwartet und konnten empirisch anhand der PIAAC-Daten belegt werden. Vorteile von Personen mit einem hohen Bildungsabschluss konnten durch höhere Lesekompetenzen erklärt werden. Die Testwerte des technologiebasierten Problemlösens bilden entstehungsgeschichtlich begründete Unabhängigkeiten des Kompetenzkonstruktes ab, das zumeist nicht schulisch erworben wurde.
4. Technologiebasiertes Problemlösen ist ein eigenständiges Kompetenzkonstrukt, das Parallelen zu mathematischen und Lesekompetenzen aufweist.
Mathematische und Lesekompetenzen stehen in einem Zusammenhang mit dem Umgang mit IKT und somit auch mit dem technologiebasierten Problemlösen. Die Leseanteile am technologiebasierten Problemlösen werden als größer eingeschätzt, als die der mathematischen Kompetenz und stärkere Effekte durch Lesekompetenz konnten empirisch
nachgewiesen werden. Die Annahme der Eigenständigkeit des technologiebasierten Problemlösens wird im Weiteren dadurch gestützt, dass der alltägliche Umgang mit mathematischen und Leseinhalten weniger stark mit ihm assoziiert ist, als der Umgang mit IKT. Nomologische Netze zwischen technologiebasiertem Problemlösen und mathematischen sowie Lesekompetenzen konnten empirisch gestützt werden und die Annahme der Eigenständigkeit des Kompetenzkonstruktes bestärkt.
5. Die Nutzung von IKT – als Lerngelegenheit verstanden – und andere Lerngelegenheiten sowie Indikatoren erfolgreichen Lernens können zum technologiebasierten Problemlösen beitragen.
Lebenslanges Lernen ist ein Teil der menschlichen Natur und in diesem Rahmen dienen verschiedenste formelle und informelle Lerngelegenheiten dem Erwerb von Wissen und Kompetenzen (Dohmen, 2001). So erreichen Personen im technologiebasierten Problemlösen ein höheres Kompetenzniveau, wenn sie regelmäßig privaten oder beruflichen Umgang mit IKT hatten. Neben diesen informellen Lerngelegenheiten steht auch die Teilnahme an Weiterbildungen in einem positiven Zusammenhang mit der technologiebasierten Bewältigung von Problemen. Des Weiteren hat eine positive Einstellung gegenüber dem Lernen neuer Inhalte einen Einfluss auf die technologiebasierte Problemlösekompetenz. Die Testwerte des technologiebasierten Problemlösens spiegeln die Lernerfolge durch das lebenslange Lernen wider, die durch verschiedene bildungsbiographische Merkmale – insbesondere formelle und informelle Lerngelegenheiten – befördert werden.
Das technologiebasierte Problemlösen in der Operationalisierung der PIAAC-Studie lässt eine konstruktrepräsentative Testwertinterpretation zu und ermöglicht somit eine differenzierte Beschreibung von Kompetenzen im Umgang mit IKT.
Die vorliegende Arbeit stellt ein organisches Taskverarbeitungssystem vor, das die zuverlässige Verwaltung und Verarbeitung von Tasks auf Multi-Core basierten SoC-Architekturen umsetzt. Aufgrund der zunehmenden Integrationsdichte treten bei der planaren Halbleiter-Fertigung vermehrt Nebeneffekte auf, die im Systembetrieb zu Fehler und Ausfällen von Komponenten führen, was die Zuverlässigkeit der SoCs zunehmend beeinträchtigt. Bereits ab einer Fertigungsgröße von weniger als 100 nm ist eine drastische Zunahme von Elektromigration und der Strahlungssensitivität zu beobachten. Gleichzeitig nimmt die Komplexität (Applikations-Anforderungen) weiter zu, wobei der aktuelle Trend auf eine immer stärkere Vernetzung von Geräten abzielt (Ubiquitäre Systeme). Um diese Herausforderungen autonom bewältigen zu können, wird in dieser Arbeit ein biologisch inspiriertes Systemkonzept vorgestellt. Dieses bedient sich der Eigenschaften und Techniken des menschlichen endokrinen Hormonsystems und setzt ein vollständig dezentrales Funktionsprinzip mit Selbst-X Eigenschaften aus dem Organic Computing Bereich um. Die Durchführung dieses organischen Funktionsprinzips erfolgt in zwei getrennten Regelkreisen, die gemeinsam die dezentrale Verwaltung und Verarbeitung von Tasks übernehmen. Der erste Regelkreis wird durch das künstliche Hormonsystem (KHS) abgebildet und führt die Verteilung aller Tasks auf die verfügbaren Kerne durch. Die Verteilung erfolgt durch das Mitwirken aller Kerne und berücksichtigt deren lokale Eignung und aktueller Zustand. Anschließend erfolgt die Synchronisation mit dem zweiten Regelkreis, der durch die hormongeregelte Taskverarbeitung (HTV) abgebildet wird und einen dynamischen Task-Transfer gemäß der aktuellen Verteilung vollzieht. Dabei werden auch die im Netz verfügbaren Zustände von Tasks berücksichtigt und es entsteht ein vollständiger Verarbeitungspfad, ausgehend von der initialen Taskzuordnung, hinweg über den Transfer der Taskkomponenten, gefolgt von der Erzeugung der lokalen Taskinstanz bis zum Start des zugehörigen Taskprozesses auf dem jeweiligen Kern. Die System-Implementierung setzt sich aus modularen Hardware- und Software-Komponenten zusammen. Dadurch kann das System entweder vollständig in Hardware, Software oder in hybrider Form betrieben und genutzt werden. Mittels eines FPGA-basierten Prototyps konnten die formal bewiesenen Zeitschranken durch Messungen in realer Systemumgebung bestätigt werden. Die Messergebnisse zeigen herausragende Zeitschranken bezüglich der Selbst-X Eigenschaften. Des Weiteren zeigt der quantitative Vergleich gegenüber anderen Systemen, dass der hier gewählte dezentrale Regelungsansatz bezüglich Ausfallsicherheit, Flächen- und Rechenaufwand deutlich überlegen ist.
In dieser Arbeit werden Verfahren vorgestellt, mit dem sich hochaufgelöste wissenschaftliche Illustrationen in einem interaktiven Vorgang erstellen lassen. Die Basis dafür bildet die neu eingeführte GPU-basierte Illustrations-Pipeline, in der auf Grundlage eines 3D-Modells Bildebenen frei angelegt und miteinander kombiniert werden können. In einer Ebene wird ein bestimmter Aspekt der Illustration mit einer auswählbaren Technik gezeigt. Die Parameter der Technik sind interaktiv editierbar. Um Effizienz zu gewährleisten ist das gesamte Verfahren so konzipiert, dass es soweit wie möglich die Berechnungen auf der GPU durchführt. So ist es möglich, dass die Illustrationen mit interaktiven Frameraten gerendert werden.
The number of multilingual texts in the World Wide Web (WWW) is increasing dramatically and a multilingual economic zone like the European Union (EU) requires the availability of multilingual Natural Language Processing (NLP) tools. Due to a rapid development of NLP tools, many lexical, syntactic, semantic and other linguistic features have been used in different NLP applications. However, there are some situations where these features can not be used due the application type or unavailability of NLP resources for some of the languages. That is why an application that is intended to handle multilingual texts must have features that are not dependent on a particular language and specific linguistic tools. In this thesis, we will focus on two such applications: text readability and source and translation classification.
In this thesis, we provide 18 features that are not only suitable for both applications, but are also language and linguistic tools independent. In order to build a readability classifier, we use texts from three different languages: English, German and Bangla. Our proposed features achieve a classification accuracy that is comparable with a classifier using 40 linguistic features. The readability classifier achieves a classification F-score of 74.21% on the English Wikipedia corpus, an F-score of 75.47% on the English textbook corpus, an F-score of 86.46% on the Bangla textbook corpus and an F-score of 86.26% on the German GEO/GEOLino corpus.
We used more than two million sentence pairs from 21 European languages in order to build the source and translation classifier. The classifier using the same eighteen features achieves a classification accuracy of 86.63%. We also used the same features to build a classifier that classifies translated texts based on their origin. The classifier achieves classification accuracy of 75% for texts from 10 European languages. In this thesis, we also provide four different corpora, three for text readability analysis and one for corpus based translation studies.
Sparse coding is a popular approach to model natural images but has faced two main challenges: modelling low-level image components (such as edge-like structures and their occlusions) and modelling varying pixel intensities. Traditionally, images are modelled as a sparse linear superposition of dictionary elements, where the probabilistic view of this problem is that the coefficients follow a Laplace or Cauchy prior distribution. We propose a novel model that instead uses a spike-and-slab prior and nonlinear combination of components. With the prior, our model can easily represent exact zeros for e.g. the absence of an image component, such as an edge, and a distribution over non-zero pixel intensities. With the nonlinearity (the nonlinear max combination rule), the idea is to target occlusions; dictionary elements correspond to image components that can occlude each other. There are major consequences of the model assumptions made by both (non)linear approaches, thus the main goal of this paper is to isolate and highlight differences between them. Parameter optimization is analytically and computationally intractable in our model, thus as a main contribution we design an exact Gibbs sampler for efficient inference which we can apply to higher dimensional data using latent variable preselection. Results on natural and artificial occlusion-rich data with controlled forms of sparse structure show that our model can extract a sparse set of edge-like components that closely match the generating process, which we refer to as interpretable components. Furthermore, the sparseness of the solution closely follows the ground-truth number of components/edges in the images. The linear model did not learn such edge-like components with any level of sparsity. This suggests that our model can adaptively well-approximate and characterize the meaningful generation process.
FIAS Scientific Report 2014
(2015)
FIAS Scientific Report 2013
(2014)
FIAS Scientific Report 2012
(2013)
Acceleration of Biomedical Image Processing and Reconstruction with FPGAs
Increasing chip sizes and better programming tools have made it possible to increase the boundaries of application acceleration with reconfigurable computer chips. In this thesis the potential of acceleration with Field Programmable Gate Arrays (FPGAs) is examined for applications that perform biomedical image processing and reconstruction. The dataflow paradigm was used to port the analysis of image data for localization microscopy and for 3D electron tomography from an imperative description towards the FPGA for the first time.
After the primitives of image processing on FPGAs are presented, a general workflow is given for analyzing imperative source code and converting it to a hardware pipeline where every node processes image data in parallel. The theoretical foundation is then used to accelerate both example applications. For localization microscopy, an acceleration of 185 compared to an Intel i5 450 CPU was achieved, and electron tomography could be sped up by a factor of 5 over an Nvidia Tesla C1060 graphics card while maintaining full accuracy in both cases.
Im World Wide Web werden diverse Dinge kostenlos angeboten. So auch die so genannte Open-Source-Software. Dass viele Akteure freie und kostenlos zugängliche Software entwickeln, ohne in geregelter Weise dafür entlohnt zu werden, wirft die Frage nach der grundsätzlichen Vereinbarkeit dieses Phänomens mit dem herrschenden kapitalistischen Wirtschaftssystem auf. Ökonomisches Verhalten scheint außer Kraft gesetzt. Statt von „Ökonomie“ könnte man in diesem Fall von „Geschenkökonomie“ bzw. „Gift Economy“ reden. Es stellt sich die Frage, welche Rolle Gaben/Geschenke im Rahmen des Open-Source-Phänomens spielen und inwiefern dieses Phänomen durch die vorhandenen Theorien zur Gift Economy erklärt werden kann. Um die Forschungsfragen zufrieden stellend zu beantworten, bedarf es einerseits einer Präzisierung auf theoretischer Ebene, die neben den bestehenden Theorien zur Gift Economy das Phänomen Open-Source im Besonderen berücksichtigt. Darüber hinaus soll das Phänomen Open Source in Form einer qualitativen empirischen Studie, welche die je spezifischen Motive des ,Zusammenhandelns’ der Open-Source-Programmierer zum Gegenstand hat, daraufhin untersucht werden, welche Rolle Gabe/Geschenke in Bezug auf dieses spielen. Ziel ist es hierbei, eine Typologie des Open-Source-Programmierers zu entwickeln. Anhand dieser Typologie soll deutlich werden, welche unterschiedlichen thematischen Kontexte für das Phänomen Open-Source rele¬vant sind. Auf Basis der Forschungsergebnisse ist es dann zudem möglich, in einer Schlussfolgerung die Theorie der Gift Economy auf der sicheren Basis em¬pirischer Daten gegebenenfalls weitergehend zu differenzieren.
Ein Ansatz für semantisches Selbstmanagement von verteilten Anwendungen im privaten Lebensumfeld
(2014)
Die Anreicherung des privaten Lebensumfelds mit intelligenten technischen Assistenzsystemen wird in den nächsten Jahrzehnten stark zunehmen. Als Teil dieser Entwicklung wird die Nutzung von externen und hauseigenen IT-Diensten steigen, wodurch sich auch die Komplexität der entstehenden Gesamtsysteme erhöht. Hier sind Ansätze gefordert, diese Systeme auch für technisch nicht versierte Benutzer produktiv nutzbar und beherrschbar zu gestalten, um eine Überforderung zu vermeiden. Im Umfeld häuslicher Dienstplattformen, die eine zentrale Rolle in solchen Systemen übernehmen, nimmt seit ein paar Jahren die Bedeutung der semantischen Modellierung von Diensten stark zu. Diese dient zum einen der formalen Repräsentation von zugehörigen Kontextinformationen, die durch Interaktion mit Sensoren und Aktoren entstehen, und zum anderen der Verbesserung der Interoperabilität zwischen Systemen unterschiedlicher Hersteller. Bisherige Ansätze beschränken sich jedoch auf den Einsatz eines zentralen Rechenknotens zur Ausführung der Dienstplattform und nutzen Semantik – wenn überhaupt – nur zur Verarbeitung von Kontextinformationen. Ein technisches Management des Gesamtsystems findet i.d.R. nicht statt.
Vor diesem Hintergrund ist das Ziel dieser Arbeit die Entwicklung eines Ansatzes für semantisches Selbstmanagement von verteilten dienstbasierten Anwendungen speziell im Umfeld häuslicher Dienstplattformen.
Die vorliegende Arbeit definiert zunächst formale Ontologien für Dienste, Dienstgütemanagement, Selbstmanagement und zugehörige Managementregeln, die zur Laufzeit mit konkreten Diensten und deren erfassten Leistungskenngrößen integriert werden. Durch einen modellgetriebenen Architekturansatz (Model Driven Architecture, MDA) wird ein technologieunabhängiges Management auf abstrakter Ebene ermöglicht, das die Wiederverwendbarkeit von Managementregeln in anderen Szenarien erlaubt.
Dieser Ansatz wird zunächst in eine Architektur für einen hochverfügbaren autonomen Manager überführt, der die Überwachung und Steuerung von Diensten und zugehörigen Dienstplattformen übernehmen kann und auf der aus dem Autonomic Computing bekannten MAPE-K-Kontrollschleife (Monitor, Analyze, Plan, Execute, Knowledge) basiert.
Den Abschluss der Arbeit bildet eine qualitative und quantitative Evaluation (mittels einer OSGi-basierten prototypischen Umsetzung) der erreichten Ergebnisse, die einen Einsatz über die Grenzen des privaten Lebensumfelds hinaus nahelegen.
On development, feasibility, and limits of highly efficient CPU and GPU programs in several fields
(2013)
With processor clock speeds having stagnated, parallel computing architectures have achieved a breakthrough in recent years. Emerging many-core processors like graphics cards run hundreds of threads in parallel and vector instructions are experiencing a revival. Parallel processors with many independent but simple arithmetical logical units fail executing serial tasks efficiently. However, their sheer parallel processing power makes them predestined for parallel applications while the simple construction of their cores makes them unbeatably power efficient. Unfortunately, old programs cannot profit by simple recompilation. Adaptation often requires rethinking and modifying algorithms to make use of parallel execution. Many applications have some serial subroutines which are very hard to parallelize, hence contemporary compute clusters are often homogeneous, offering fast processors for serial tasks and parallel processors for parallel tasks. In order not to waste the available compute power, highly efficient programs are mandatory.
This thesis is about the development of fast algorithms and their implementations on modern CPUs and GPUs, about the maximum achievable efficiency with respect to peak performance and to power consumption respectively, and about feasibility and limits of programs for CPUs, GPUs, and heterogeneous systems. Three totally different applications from distinct fields, which were developed in the extent of this thesis, are presented.
The ALICE experiment at the LHC particle collider at CERN studies heavy-ion collisions at high rates of several hundred Hz, while every collision produces thousands of particles, whose trajectories must be reconstructed. For this purpose, ALICE track reconstruction and ALICE track merging have been adapted for GPUs and deployed on 64 GPU-enabled compute-nodes at CERN.
After a testing phase, the tracker ran in nonstop operation during 2012 providing full real-time track reconstruction. The tracker employs a multithreaded pipeline as well as asynchronous data transfer to ensure continuous GPU utilization and outperforms the fastest available CPUs by about a factor three.
The Linpack benchmark is the standard tool for ranking compute clusters. It solves a dense system of linear equations using primarily matrix multiplication facilitated by a routine called DGEMM. A heterogeneous GPU-enabled version of DGEMM and Linpack has been developed, which can utilize the CAL, CUDA, and OpenCL APIs as backend. Employing this implementation, the LOEWE-CSC cluster ranked place 22 in the November 2010 Top500 list of the fastest supercomputers, and the Sanam cluster achieved the second place in the November 2012 Green500 list of the most power efficient supercomputers. An elaborate lookahead algorithm, a pipeline, and asynchronous data transfer hide the serial CPU-bound tasks of Linpack behind DGEMM execution on the GPU reaching the highest efficiency on GPU-accelerated clusters.
Failure erasure codes enable failure tolerant storage of data and real-time failover, ensuring that in case of a hardware defect servers and even complete data centers remain operational. It is an absolute necessity for present-day computer infrastructure. The mathematical theory behind the codes involves matrix-computations in finite fields, which are not natively supported by modern processors and hence computationally very expensive. This thesis presents a novel scheme for fast encoding matrix generation and demonstrates a fast implementation for the encoding itself, which uses exclusively either integer or logical vector instructions. Depending on the scenario, it is always hitting different hard limits of the hardware: either the maximum attainable memory bandwidth, or the peak instruction throughput, or the PCI Express bandwidth limit when GPUs or FPGAs are used.
The thesis demonstrates that in most cases with respect to the available peak performance, GPU implementations can be as efficient as their CPU counterparts.
With respect to costs or power consumption, they are much more efficient. For this purpose, complex tasks must be split in serial as well as parallel parts and the execution must be pipelined such that the CPU bound tasks are hidden behind GPU execution. Few cases are identified where this is not possible due to PCI Express limitations or not reasonable because practical GPU languages are missing.
The human brain is an unparalleled system: Through millions of years of evolution and during a lifespan of learning, our brains have developed remarkable abilities for dealing with incoming sensory data, extracting structure and useful information, and finally drawing the conclusions that result in the actions we take. Understanding the principles behind this machinery and building artificial systems that mimic at least some of these capabilities is a long standing goal in both the scientific and the engineering communities. While this goal still seems unreachable, we have seen tremendous progress when it comes to training data-driven algorithms on vast amounts of training data, e.g. to learn an optimal data model and its parameters in order to accomplish some task. Such algorithms are now omnipresent: they are part of recommender systems, they perform speech recognition and generally build the foundation for many semi-autonomous systems. They start to be integral part of many technical systems modern technical societies rely on for their everyday functioning. Many of these algorithms were originally inspired by biological systems or act as models for sensory data processing in mammalian brains. The response properties of a certain population of neurons in the first stages of the mammalian visual pathway, for example, can be modeled by algorithms such as Sparse Coding (SC), Independent Component Analysis (ICA) or Factor Analysis (FA). These well established learning algorithms typically assume linear interactions between the variables of the model. Most often these relationships are expressed in the form of a matrix-vector products between a matrix with learned dictionary-elements (basis vectors as column vectors) and the latent variables of these models. While on the one hand this linear interaction can sometimes be justified by the physical process for which the machine learning model is proposed, it is on the other hand often chosen just because of its mathematical and practical convenience. From an optimal coding point of view though, one would generally expect that the ideal model closely reflect the core interactions of the system it is modeling. In vision for example, one of the dominant processes giving rise to our sensory percepts are occlusions. Occluding objects are omnipresent in visual scenes and it would not be surprising if the mammalian visual system would be optimized to process occluding structures in the visual data stream. Yet, the established mathematical models of the first stages of the visual processing path (like, e.g., SC, ICA or FA) all assume linear interactions between the active image components. In this thesis we will discuss new models that aim to approximate the effects of occluding components by assuming nonlinear interactions between their activated dictionary elements. We will present learning algorithms that infer optimal parameters for these models given data. In the experiments, we will validate the algorithms on artificial ground truth data and demonstrate their ability to recover the correct model parameters. We will show that the predictions made by these nonlinear models correspond better to the experimental data measured in-vivo than the predictions made by the established linear models. Furthermore, we systematically explore and compare a large space of plausible combinations of hyperparameters and preprocessing schemes in order to eliminate any effects of artefacts on the observed results. Training nonlinear sparse coding models is computationally more demanding than training linear models. In order to perform the numerical experiments described in this thesis we developed a software framework that facilitates the implementation of massive parallel expectation maximization (EM) based learning algorithms. This infrastructure was used for all experiments described in here, as well as by collaborators in projects we will not discuss. Some of the experiments required more than 1017 floating point operations and were run on a computer cluster running on up to 5000 CPU Cores in parallel. Our parallel framework enabled these experiments to be performed.
Driven by rapid technological advancements, the amount of data that is created, captured, communicated, and stored worldwide has grown exponentially over the past decades. Along with this development it has become critical for many disciplines of science and business to being able to gather and analyze large amounts of data. The sheer volume of the data often exceeds the capabilities of classical storage systems, with the result that current large-scale storage systems are highly distributed and are comprised of a high number of individual storage components. As with any other electronic device, the reliability of storage hardware is governed by certain probability distributions, which in turn are influenced by the physical processes utilized to store the information. The traditional way to deal with the inherent unreliability of combined storage systems is to replicate the data several times. Another popular approach to achieve failure tolerance is to calculate the block-wise parity in one or more dimensions. With better understanding of the different failure modes of storage components, it has become evident that sophisticated high-level error detection and correction techniques are indispensable for the ever-growing distributed systems. The utilization of powerful cyclic error-correcting codes, however, comes with a high computational penalty, since the required operations over finite fields do not map very well onto current commodity processors. This thesis introduces a versatile coding scheme with fully adjustable fault-tolerance that is tailored specifically to modern processor architectures. To reduce stress on the memory subsystem the conventional table-based algorithm for multiplication over finite fields has been replaced with a polynomial version. This arithmetically intense algorithm is better suited to the wide SIMD units of the currently available general purpose processors, but also displays significant benefits when used with modern many-core accelerator devices (for instance the popular general purpose graphics processing units). A CPU implementation using SSE and a GPU version using CUDA are presented. The performance of the multiplication depends on the distribution of the polynomial coefficients in the finite field elements. This property has been used to create suitable matrices that generate a linear systematic erasure-correcting code which shows a significantly increased multiplication performance for the relevant matrix elements. Several approaches to obtain the optimized generator matrices are elaborated and their implications are discussed. A Monte-Carlo-based construction method allows it to influence the specific shape of the generator matrices and thus to adapt them to special storage and archiving workloads. Extensive benchmarks on CPU and GPU demonstrate the superior performance and the future application scenarios of this novel erasure-resilient coding scheme.
Paging is one of the most prominent problems in the field of online algorithms. We have to serve a sequence of page requests using a cache that can hold up to k pages. If the currently requested page is in cache we have a cache hit, otherwise we say that a cache miss occurs, and the requested page needs to be loaded into the cache. The goal is to minimize the number of cache misses by providing a good page-replacement strategy. This problem is part of memory-management when data is stored in a two-level memory hierarchy, more precisely a small and fast memory (cache) and a slow but large memory (disk). The most important application area is the virtual memory management of operating systems. Accessed pages are either already in the RAM or need to be loaded from the hard disk into the RAM using expensive I/O. The time needed to access the RAM is insignificant compared to an I/O operation which takes several milliseconds.
The traditional evaluation framework for online algorithms is competitive analysis where the online algorithm is compared to the optimal offline solution. A shortcoming of competitive analysis consists of its too pessimistic worst-case guarantees. For example LRU has a theoretical competitive ratio of k but in practice this ratio rarely exceeds the value 4.
Reducing the gap between theory and practice has been a hot research issue during the last years. More recent evaluation models have been used to prove that LRU is an optimal online algorithm or part of a class of optimal algorithms respectively, which was motivated by the assumption that LRU is one of the best algorithms in practice. Most of the newer models make LRU-friendly assumptions regarding the input, thus not leaving much room for new algorithms.
Only few works in the field of online paging have introduced new algorithms which can compete with LRU as regards the small number of cache misses.
In the first part of this thesis we study strongly competitive randomized paging algorithms, i.e. algorithms with optimal competitive guarantees. Although the tight bound for the competitive ratio has been known for decades, current algorithms matching this bound are complex and have high running times and memory requirements. We propose the algorithm OnlineMin which processes a page request in O(log k/log log k) time in the worst case. The best previously known solution requires O(k^2) time.
Usually the memory requirement of a paging algorithm is measured by the maximum number of pages that the algorithm keeps track of. Any algorithm stores information about the k pages in the cache. In addition it can also store information about pages not in cache, denoted bookmarks. We answer the open question of Bein et al. '07 whether strongly competitive randomized paging algorithms using only o(k) bookmarks exist or not. To do so we modify the Partition algorithm of McGeoch and Sleator '85 which has an unbounded bookmark complexity, and obtain Partition2 which uses O(k/log k) bookmarks.
In the second part we extract ideas from theoretical analysis of randomized paging algorithms in order to design deterministic algorithms that perform well in practice. We refine competitive analysis by introducing the attack rate
parameter r, which ranges between 1 and k. We show that r is a tight bound on the competitive ratio of deterministic algorithms.
We give empirical evidence that r is usually much smaller than k and thus r-competitive algorithms have a reasonable performance on real-world traces. By introducing the r-competitive priority-based algorithm class OnOPT we obtain a collection of promising algorithms to beat the LRU-standard. We single out the new algorithm RDM and show that it outperforms LRU and some of its variants on a wide range of real-world traces.
Since RDM is more complex than LRU one may think at first sight that the gain in terms of lowering the number of cache misses is ruined by high runtime for processing pages. We engineer a fast implementation of RDM, and compare it
to LRU and the very fast FIFO algorithm in an overall evaluation scheme, where we measure the runtime of the algorithms and add penalties for each cache miss.
Experimental results show that for realistic penalties RDM still outperforms these two algorithms even if we grant the competitors an idealistic runtime of 0.
Two generic mechanisms for emergence of direction selectivity coexist in recurrent neural networks
(2013)
Poster presentation: Twenty Second Annual Computational Neuroscience Meeting: CNS*2013. Paris, France. 13-18 July 2013.
In the mammalian visual cortex, the time-averaged response of many neurons is maximal for stimuli moving in a particular direction. Such a direction selective response is not found in LGN, upstream of the visual processing pathway, suggesting that cortical networks play a strong role in the generation of direction selectivity. Here we investigate the mechanisms for the emergence of direction selectivity in the recurrent networks of nonlinear firing rate neurons in layer 4 of V1 receiving the input from LGN. In the model the LGN inputs are characterized by different receptive field positions, and their relative temporal phase shifts are reversed for the stimuli moving in the opposite direction. We propose that two distinct mechanisms result in the neuronal direction selective response in these recurrent networks. The first one is a result of nonlinear feed-forward summation of several time-shifted inputs. The second mechanism is based on the competition between neurons for firing in a winner-take-all regime. Both mechanisms rely on inhibitory interactions in the connectivity matrix of lateral connections, but the second one involves inhibitory loops. Typically, the first mechanism results in lower selectivity values than the second, but the time-course of acquiring direction selective response is faster for the first mechanism. Importantly, the two mechanisms have different input frequency tuning. The first mechanism, based on the nonlinear summation, result in a relatively narrow tuning curve around the preferred frequency of the stimulus in the case of the moving grating. In contrast, the direction selectivity arising from the second mechanism depends only weakly on the input frequency, i.e. has a broader tuning curve. These differences allow us to provide the recipe for identifying in experiment which of the two mechanisms is used by a given direction selective neuron. We then analyze how the statistics of the connections in the random recurrent networks affect the relative contributions from these two mechanisms and determine the distributions of the direction selectivity values. We identify the motifs in the connectivity matrix, which are required for each mechanism and show that the minimal conditions for both mechanisms are met in a very broad set of random recurrent networks with sufficiently strong inhibitory connections. Thus, we propose that these mechanisms coexist in generic recurrent networks with inhibition. Our results may account for the recent experimental observations that direction selectivity is present in dark-reared mice and ferrets [1,2]. It can also explain the emergence of direction selectivity in species lacking a spatially organized direction selectivity map.
Finding motifs in biological, social, technological, and other types of networks has become a widespread method to gain more knowledge about these networks’ structure and function. However, this task is very computationally demanding, because it is highly associated with the graph isomorphism which is an NP problem (not known to belong to P or NP-complete subsets yet). Accordingly, this research is endeavoring to decrease the need to call NAUTY isomorphism detection method, which is the most time-consuming step in many existing algorithms. The work provides an extremely fast motif detection algorithm called QuateXelero, which has a Quaternary Tree data structure in the heart. The proposed algorithm is based on the well-known ESU (FANMOD) motif detection algorithm. The results of experiments on some standard model networks approve the overal superiority of the proposed algorithm, namely QuateXelero, compared with two of the fastest existing algorithms, G-Tries and Kavosh. QuateXelero is especially fastest in constructing the central data structure of the algorithm from scratch based on the input network.
Das Ziel dieser Arbeit ist es, eine authentische Verdeckung eingebetteter virtueller 3D-Objekte in augmentierten Bilderwelten bei einer geringen Anzahl an Fotos innerhalb der Bilderwelt zu erreichen. Für die Verdeckung von realen und virtuellen Anteilen einer Augmented Reality-Szene sind Tiefeninformationen notwendig. Diese stammen üblicherweise aus einer 3D-Rekonstruktion, für deren Erstellung sehr viele Eingangsbilder notwendig sind. Im Gegensatz dazu wurde in dieser Arbeit ein System entwickelt, das eine vollständige 3D-Rekonstruktion umgeht. Dieses beruht auf einem direkten bildbasierten Rendering-Ansatz, welcher auch mit unvollständigen Tiefeninformationen eine hohe Bildqualität in Bezug auf eine authentische Verdeckung erreicht. Daraus erschließen sich neue Anwendungsgebiete, wie z.B. die automatisierte Visualisierung von 3D-Planungsdaten und 3D-Produktpräsentationen in Bildern bzw. Bilderwelten, da in diesen Bereichen oftmals nicht genügend große Bildmengen vorhanden sind. Gerade für diese Anwendungsgebiete sind authentische Verdeckungen für die Nutzerakzeptanz der Augmentierung wichtig. Unter authentischer Verdeckung wird die entsprechend der menschlichen Wahrnehmung visuell korrekte Überlagerung zwischen virtuellen Objekten und einzelnen Bildanteilen eines oder mehrerer Fotos verstanden. Das Ergebnis wird in Form einer Bilderwelt (eine bildbasierte 3D-Welt, die die Fotos entsprechend der Bildinhalte räumlich anordnet) präsentiert, die mit virtuellen Objekten erweitert wurde. Folglich ordnet sich diese Arbeit in das Fachgebiet der Augmented Reality ein. Im Rahmen dieser Arbeit wurde ein Verfahren für die bildbasierte Darstellung mit authentischen Verdeckungen auf der Basis von unvollständigen Tiefeninformationen sowie unterschiedliche Verfahren für die notwendige Berechnung der Tiefeninformationen entwickelt und gegenübergestellt. Das Sliced-Image-Rendering-Verfahren rendert mithilfe unvollständiger Tiefeninformationen ein Bild ohne 3D-Geometrie als dreidimensionale Darstellung und realisiert auf diese Weise eine authentische Verdeckung. Das Berechnen der dafür notwendigen Tiefeninformationen eines 2D-Bildes stellt eine gesonderte Herausforderung dar, da die Bilderwelt nur wenige und unvollständige 3D-Informationen der abgebildeten Szene bereitstellt. Folglich kann eine qualitativ hochwertige 3D-Rekonstruktion nicht durchgeführt werden. Die Fragestellung ist daher, wie einzelne Tiefeninformationen berechnet und diese anschließend größeren Bildbereichen zugeordnet werden können. Für diese Tiefenzuordnung wurden im Rahmen der vorliegenden Arbeit drei verschiedene Verfahren konzipiert, die sich in Bezug auf genutzte Daten und deren Verarbeitung unterscheiden. Das Segment-Depth-Matching-Verfahren ordnet Segmenten eines Bildes mithilfe der 3D-Szeneninformationen der Bilderwelt eine Tiefe zu. Hierfür werden Segmentbilder vorausgesetzt. Als Ergebnis liegt für jedes Foto eine Depth-Map vor. Um eine Tiefenzuordnung auch ohne eine vorangehende Segmentierung zu ermöglichen, wurde das Key-Point-Depth-Matching-Verfahren entwickelt. Bei diesem Verfahren werden die 3D-Szeneninformationen der Bilderwelt auf die Bildebene als kreisförmige Sprites projiziert. Die Distanz zur Kamera wird dabei als Tiefenwert für das Sprite verwendet. Alle projizierten Sprites einer Kamera ergeben die Depth-Map. Beide Verfahren liefern Flächen mit Tiefeninformationen, aber keine pixelgenauen Depth-Maps. Um pixelgenaue Depth-Maps zu erzeugen, wurde das Geometry-Depth-Matching-Verfahren entwickelt. Bei diesem Verfahren wird eine Szenengeometrie des abgebildeten Szenenausschnittes erzeugt und dadurch eine pixelgenaue Depth-Map erstellt. Hierfür wird ein semiautomatischer Skizzierungsschritt vorausgesetzt. Die erzeugte Szenengeometrie stellt keine vollständige 3D-Rekonstruktion der Bilderweltenszene dar, da nur ein Szenenausschnitt aus Sicht einer Kamera rekonstruiert wird. Anhand einer technischen Umsetzung erfolgte eine Validierung der konzeptionellen Verfahren. Die daraus resultierenden Ergebnisse wurden anhand verschiedener Bilderweltenszenen mit unterschiedlichen Eigenschaften (Außen- und Innenraumszenen, detailreich und -arm, unterschiedliche Bildmengen) evaluiert. Die Evaluierung des Sliced-Image-Renderings zeigt, dass mithilfe unvollständiger Tiefeninformationen der entwickelten Depth-Matching-Verfahren und unter Einhaltung der gestellten Anforderungen (wenig Eingabefotos, kleine Szenen, keine 3D-Rekonstruktion) eine authentische Verdeckung eingebetteter virtueller 3D-Objekte in Bilderwelten realisiert werden kann. Mithilfe des entwickelten Systems können bildbasierte Anwendungen auch mit kleinen Fotomengen Augmentierungen mit hoher Bildqualität in Bezug auf eine authentische Verdeckung realisieren.
In der modernen Hochschullehre haben sich eLearning-Elemente als ein Teil des Lehrrepertoires etabliert. Der Einsatz interaktiver webbasierter Selbstlernmodule (Web Based Trainings (WBT)) ist dabei eine Option. Hochschulen und Unternehmen versprechen sich dadurch neue Möglichkeiten des Lehrens und Lernens, um z. B. einen Ausgleich heterogener Vorerfahrungen sowie eine stärkere aktive Beteiligung der Lernenden zu bewirken. Damit die Erstellung und Strukturierung dieser Inhalte mit möglichst geringem Aufwand erfolgen kann, bieten Autorensysteme Unterstützung.
Zu den Grundfunktionen von Autorensystemen gehören unter anderem, das Einbinden gebräuchlicher Medienformate, die einfache Erstellung von Fragen sowie verschiedene Auswertungs- und Feedbackmöglichkeiten. Obwohl Autorensysteme schon vor vielen Jahren ihre erste praktische Anwendung fanden, gibt es nach wie vor Schwachstellen, die sich auf den gesamten Erstellungsprozess wie auch auf einzelne Funktionen beziehen. Im Detail wird bemängelt, dass die Werkzeuge zu komplex und unflexibel sind. Darüber hinaus fehlt häufig eine zufriedenstellende Verknüpfung der vielen Werkzeuge entlang der Prozesskette zu einer Gesamtlösung.
Des Weiteren wird die Konzentration auf die Produktionsphase kritisiert, wodurch andere wichtige Prozesse in den Hintergrund treten bzw. außer Acht gelassen werden.
Im Rahmen der Zusammenarbeit mit einem Automobilhersteller, für den die erste Version des Autorensystems LernBar weiterentwickelt wurde, spielte der Begriff „Lean Production“ inhaltlich in der Umsetzung der WBTs eine wesentliche Rolle. Die Lean Production, die über viele Jahre für die Automobilindustrie entwickelt, verbessert und angepasst wurde, liefert Optimierungsansätze für den Produktionsbereich. Ein wirtschaftlicher Nutzen des Lean-Ansatzes wird auch in anderen Bereichen gesehen wie z. B. in der Softwareentwicklung („Lean Software Development“) oder im Management („Lean Management“). Dabei bietet die Wertschöpfungsorientierung Lösungen für die widersprüchlichen Ziele mehr Leistungen zu geringeren Kosten, schneller und in höherer Qualität zugleich zu liefern. Aus der Grundidee der Lean Production entwickelte sich vorliegendes Dissertationsthema in Bezug darauf, inwiefern sich diese Prinzipien auf den WBT-Produktionsprozess übertragen lassen und die LernBar (das hierfür weiterentwickelnde Autorensystem) dabei Unterstützung bieten kann.
Zunächst wurde analysiert, welche Werkzeuge und Hilfestellungen benötigt werden, um unter dem Aspekt der Lean Production WBTs im universitären Umfeld erstellen zu können. In diesem Zusammenhang wurden Merkmale einer „Lean Media Production“ definiert sowie konzeptionell und technisch umgesetzt. Zur Verbesserung der Prozesse flossen Ergebnisse aus empirischer und praktischer Forschung ein. Im Vergleich zu anderen Entwicklungen bei denen häufig das Hauptziel eine umfangreiche Funktionalität ist, werden u.a. folgende übertragbare Ziele bei der Umsetzung verfolgt: Verschwendung vermeiden, eine starke Einbeziehung der Kunden, Werkzeuge die nahtlos ineinandergreifen, eine hohe Flexibilität und eine stetige Qualitätsverbesserung.
Zur Erreichung dieser Zielsetzungen wurden alle Prozesse kontinuierlich verbessert, sich auf das Wesentliche und die Wertschöpfung konzentriert sowie überflüssige Schritte eliminiert. Demnach ist unter dem Begriff „Lean Media Production“ ein skalierbarer, effizienter und effektiver Produktionsprozess zu verstehen, in dem alle Werkzeuge ineinandergreifen.
Die Realisierung der „Lean Media Production“ erfolgte anhand des Autorensystems LernBar, wobei die typischen Softwareentwicklungsphasen Entwurf, Implementierung und Evaluierung mehrfach durchlaufen wurden. Ausschlaggebend dabei war, dass der „Lean“-Aspekt berücksichtigt wurde und dies somit eine neue Vorgehensweise bei der Umsetzung eines Autorensystems darstellt. Im Verlauf der Entwicklungen ergaben sich, durch eine formative Evaluation, den Einsatz in Projekten und eine empirische Begleitforschung, neue Anforderungen an das System. Ein Vergleich der zwei Produktionssysteme, Automobil vs. WBT-Produktion, zeigt und bestätigt die Erwartung, dass nicht alle Prinzipien der Lean Production übertragbar sind.
Dennoch war diese Untersuchung notwendig, da sie Denkanstöße zur Entwicklung und Optimierung des Erstellungsprozesses eines WBTs gab. Auch die Ergebnisse der abschließenden Online-Befragung ergaben, dass die Ziele der Arbeit erreicht wurden, dass aber weiterer Optimierungsbedarf besteht. Die LernBar Release 3 bietet für alle Produktionsphasen Werkzeuge an, durch die eine effektive und effiziente Erstellung von WBTs von der Idee bis zur Distribution möglich ist.
Stand noch vor fünf Jahren zu Beginn dieser Arbeit das Endprodukt bei der LernBar Entwicklung im Vordergrund, verlagerte sich durch den Einfluss dieser Dissertation der Schwerpunkt auf den gesamten Produktionsprozess. Unter Berücksichtigung der in diesem Zusammenhang entwickelten Prinzipien einer „Lean Media Production“, nehmen bspw. die Wirtschaftlichkeit und die starke Kundenorientierung während des Produktionsprozesses einen wichtigen Stellenwert ein. Dieser Ansatz ist eine neue Vorgehensweise im Bereich der Entwicklung von Autorensystemen, der seine Anerkennung und Professionalität durch die Ergebnisse des selbstentwickelten Evaluationsbogens sowie dem stetig wachsenden Einsatz in Schulen, Hochschulen und Unternehmen belegen kann.
In weiteren Forschungsarbeiten ist zu untersuchen, welche Lean Production Prinzipien zu verwenden oder anzupassen sind, wenn z. B. in größeren Teams oder mobil produziert wird. Des Weiteren sollte überprüft werden, inwieweit die Lernenden mit dem Endprodukt zufrieden sind und in ihrem Lernprozess unterstützt werden. Durch diese Forschungsarbeit wurde ein Beitrag dazu geleistet, die Lehre und Ausbildung zu optimieren, indem die Autoren/Lehrende in der Erstellung ihrer digitalen Lerninhalte im gesamten Prozess von aufeinander abgestimmten Werkzeugen unterstützt werden.
FIAS Scientific Report
(2011)
FIAS Scientific Report 2011
(2012)
FIAS Scientific Report 2010
(2011)
In the year 2010 the Frankfurt Institute for Advanced Studies has successfully continued to follow its agenda to pursue theoretical research in the natural sciences. As stipulated in its charter, FIAS closely collaborates with extramural research institutions, like the Max Planck Institute for Brain Research in Frankfurt and the GSI Helmholtz Center for Heavy Ion Research, Darmstadt and with research groups at the science departments of Goethe University. The institute also engages in the training of young researchers and the education of doctoral students. This Annual Report documents how these goals have been pursued in the year 2010. Notable events in the scientific life of the Institute will be presented, e.g., teaching activities in the framework of the Frankfurt International Graduate School for Science (FIGSS), colloquium schedules, conferences organized by FIAS, and a full bibliography of publications by authors affiliated with FIAS. The main part of the Report consists of short one-page summaries describing the scientific progress reached in individual research projects in the year 2010...
FIAS Scientific Report 2009
(2010)
In this Annual Report we present some of the ongoing activities of FIAS and of the associated graduate
school, the “Frankfurt International Graduate School for Science” (FIGSS) in the year 2009. The main part of the Report consists of a collection of short reports describing the research projects of scientists working at or associated with FIAS.
Pre-publication peer review of scientific literature in its present state suffers from a lack of evaluation validity and transparency to the community. Inspired by social networks, we propose a framework for the open exchange of post-publication evaluation to complement the current system. We first formulate a number of necessary conditions that should be met by any design dedicated to perform open scientific evaluation. To introduce our framework, we provide a basic data standard and communication protocol. We argue for the superiority of a provider-independent framework, over a few isolated implementations, which allows the collection and analysis of open evaluation content across a wide range of diverse providers like scientific journals, research institutions, social networks, publishers websites, and more. Furthermore, we describe how its technical implementation can be achieved by using existing web standards and technology. Finally, we illustrate this with a set of examples and discuss further potential.
Conceptual design of an ALICE Tier-2 centre integrated into a multi-purpose computing facility
(2012)
This thesis discusses the issues and challenges associated with the design and operation of a data analysis facility for a high-energy physics experiment at a multi-purpose computing centre. At the spotlight is a Tier-2 centre of the distributed computing model of the ALICE experiment at the Large Hadron Collider at CERN in Geneva, Switzerland. The design steps, examined in the thesis, include analysis and optimization of the I/O access patterns of the user workload, integration of the storage resources, and development of the techniques for effective system administration and operation of the facility in a shared computing environment. A number of I/O access performance issues on multiple levels of the I/O subsystem, introduced by utilization of hard disks for data storage, have been addressed by the means of exhaustive benchmarking and thorough analysis of the I/O of the user applications in the ALICE software framework. Defining the set of requirements to the storage system, describing the potential performance bottlenecks and single points of failure and examining possible ways to avoid them allows one to develop guidelines for selecting the way how to integrate the storage resources. The solution, how to preserve a specific software stack for the experiment in a shared environment, is presented along with its effects on the user workload performance. The proposal for a flexible model to deploy and operate the ALICE Tier-2 infrastructure and applications in a virtual environment through adoption of the cloud computing technology and the 'Infrastructure as Code' concept completes the thesis. Scientific software applications can be efficiently computed in a virtual environment, and there is an urgent need to adapt the infrastructure for effective usage of cloud resources.
This thesis will first introduce in more detail the Bayesian theory and its use in integrating multiple information sources. I will briefly talk about models and their relation to the dynamics of an environment, and how to combine multiple alternative models. Following that I will discuss the experimental findings on multisensory integration in humans and animals. I start with psychophysical results on various forms of tasks and setups, that show that the brain uses and combines information from multiple cues. Specifically, the discussion will focus on the finding that humans integrate this information in a way that is close to the theoretical optimal performance. Special emphasis will be put on results about the developmental aspects of cue integration, highlighting experiments that could show that children do not perform similar to the Bayesian predictions. This section also includes a short summary of experiments on how subjects handle multiple alternative environmental dynamics. I will also talk about neurobiological findings of cells receiving input from multiple receptors both in dedicated brain areas but also primary sensory areas. I will proceed with an overview of existing theories and computational models of multisensory integration. This will be followed by a discussion on reinforcement learning (RL). First I will talk about the original theory including the two different main approaches model-free and model-based reinforcement learning. The important variables will be introduced as well as different algorithmic implementations. Secondly, a short review on the mapping of those theories onto brain and behaviour will be given. I mention the most in uential papers that showed correlations between the activity in certain brain regions with RL variables, most prominently between dopaminergic neurons and temporal difference errors. I will try to motivate, why I think that this theory can help to explain the development of near-optimal cue integration in humans. The next main chapter will introduce our model that learns to solve the task of audio-visual orienting. Many of the results in this section have been published in [Weisswange et al. 2009b,Weisswange et al. 2011]. The model agent starts without any knowledge of the environment and acts based on predictions of rewards, which will be adapted according to the reward signaling the quality of the performed action. I will show that after training this model performs similarly to the prediction of a Bayesian observer. The model can also deal with more complex environments in which it has to deal with multiple possible underlying generating models (perform causal inference). In these experiments I use di#erent formulations of Bayesian observers for comparison with our model, and find that it is most similar to the fully optimal observer doing model averaging. Additional experiments using various alterations to the environment show the ability of the model to react to changes in the input statistics without explicitly representing probability distributions. I will close the chapter with a discussion on the benefits and shortcomings of the model. The thesis continues whith a report on an application of the learning algorithm introduced before to two real world cue integration tasks on a robotic head. For these tasks our system outperforms a commonly used approximation to Bayesian inference, reliability weighted averaging. The approximation is handy because of its computational simplicity, because it relies on certain assumptions that are usually controlled for in a laboratory setting, but these are often not true for real world data. This chapter is based on the paper [Karaoguz et al. 2011]. Our second modeling approach tries to address the neuronal substrates of the learning process for cue integration. I again use a reward based training scheme, but this time implemented as a modulation of synaptic plasticity mechanisms in a recurrent network of binary threshold neurons. I start the chapter with an additional introduction section to discuss recurrent networks and especially the various forms of neuronal plasticity that I will use in the model. The performance on a task similar to that of chapter 3 will be presented together with an analysis of the in uence of different plasticity mechanisms on it. Again benefits and shortcomings and the general potential of the method will be discussed. I will close the thesis with a general conclusion and some ideas about possible future work.
This paper considers the logic FOcard, i.e., first-order logic with cardinality predicates that can specify the size of a structure modulo some number. We study the expressive power of FOcard on the class of languages of ranked, finite, labelled trees with successor relations. Our first main result characterises the class of FOcard-definable tree languages in terms of algebraic closure properties of the tree languages. As it can be effectively checked whether the language of a given tree automaton satisfies these closure properties, we obtain a decidable characterisation of the class of regular tree languages definable in FOcard. Our second main result considers first-order logic with unary relations, successor relations, and two additional designated symbols < and + that must be interpreted as a linear order and its associated addition. Such a formula is called addition-invariant if, for each fixed interpretation of the unary relations and successor relations, its result is independent of the particular interpretation of < and +. We show that the FOcard-definable tree languages are exactly the regular tree languages definable in addition-invariant first-order logic. Our proof techniques involve tools from algebraic automata theory, reasoning with locality arguments, and the use of logical interpretations. We combine and extend methods developed by Benedikt and Segoufin (ACM ToCL, 2009) and Schweikardt and Segoufin (LICS, 2010).
Wenn man in der ersten Hälfte des vergangenen Jahrhunderts bereits den Begriff "Brain Drain" (Abwanderung) gekannt hätte, dann wären damit bestimmt nicht die deutschen Wissenschaftler gemeint gewesen, denn die geistige Elite zog es noch nicht in Scharen aus ihrer Heimat. Im Gegenteil! Damals folgte die internationale wissenschaftliche Elite dem Ruf nach Deutschland, weil hier weltweit herausragende Forscherpersönlichkeiten arbeiteten und lehrten. Das galt auch für die Frankfurter Universität. Namen wie Paul Ehrlich, Franz Oppenheimer oder Friedrich Dessauer stehen für hochkarätige Forschung, die ausländische Studenten und Wissenschaftler in die Mainmetropole lockte, bis das Nazi-Regime mit der Verfolgung der jüdischen Wissenschaftler dieser Blütezeit ein jähes Ende setze und viele Forscher ins Ausland – insbesondere in die USA – fliehen mussten.
L’année académique 2003-2004 a été, s’il faut le dire, celle de la sortie officielle et du lancement des activités scientifiques du Centre d’Études en Littérature Gabonaise (CELIG). La tâche première fut celle de la mise en place d’un Conseil de direction du Centre, conformément aux règles en vigueur à la Faculté des Lettres et des Sciences Humaines de l’Université Omar Bongo de Libreville. Cinq postes furent donc créés, et l’équipe se mit au travail aussitôt. Par cet acte, notre jeune Centre entrait dans le juron des centres de recherches et laboratoires de recherche reconnus par l’Université. ...
Schwerpunktthema: Das Alter. Das Verhältnis von alten zu jungen Menschen hat sich in den vergangenen 50 Jahren in den Industrienationen dramatisch verändert. Die verlängerte Lebenserwartung und abnehmende Geburtenrate führen zu einem »Ergrauen« der Gesellschaft mit einschneidenden Folgen für Sozial- und Gesundheitssysteme. Allein in Deutschland wird es im Jahr 2050 voraussichtlich zehn Millionen über 80-Jährige geben. Wie bereiten wir uns auf diese neue Situation vor? Was hält Menschen geistig und körperlich länger fit? Wie verändert sich die Situation am Arbeitsmarkt? Was tun gegen Parkinson und Alzheimer-Demenz? Welche Hilfen erhalten die Angehörigen Pflegebedürftiger? Alle diese Fragen greift die neue Ausgabe von »Forschung Frankfurt« auf. Die Alternsforschung ist ein expandierendes Forschungsfeld an der Universität Frankfurt, das interdisziplinär von der Ebene der Gene, Moleküle und Organe bis hin zum Menschen und seinem sozialen Umfeld reicht. Seit 2004 fördert das »Forum Alternswissenschaft und Alterspolitik« Forschung, Lehre und praktische Politik zum Thema Alter und Altern, unterstützt von der Universität des Dritten Lebensalters. In den Naturwissenschaften haben sich Grundlagenforscher und klinische Forscher zusammengeschlossen, um im Expertencluster »Alzheimer und Parkinson Forschung Frankfurt« gemeinsam die beiden gängigsten Erkrankungen des Alters zu bekämpfen. In der BHF-Stiftungsprofessur Gerontopsychiatrie verbinden sich medizinische und soziale Aspekte der degenerativen Alterserkrankungen. Schließlich ist auch das neu gegründete Interdisziplinäre Zentrum für Neurowissenschaften engagiert in der Erforschung von Alterskrankheiten, die das Gehirn betreffen.
ProtoSociology is an interdisciplinary journal which crosses the borders of philosophy, social sciences, and their corresponding disciplines. Each issue concentrates on a specific topic taken from the current discussion to which scientists from different fields contribute the results of their research. ProtoSociology is further a project that examines the nature of mind, language and social systems. In this context theoretical work has been done by investigating such theoretical concepts like interpretation and (social) action, globalization, the global world-system, social evolution, and the sociology of membership. Our purpose is to initiate and enforce basic research on relevant topics from different perspectives and traditions.
Rezensionen zu: Evolution : die Theorie eines selbstverständlichen Prozesses / D. Stefan Peters. Basilisken-Presse, Rangsdorf, 2010. ISBN 978-3-941365-07-0, 103 Seiten. 18 Euro Paleogene Fossil Birds / Gerald Mayr. Springer Verlag, Berlin [u.a.], 2009. ISBN 978-3-540-89627-2, 275 Seiten. 119,95 Euro
MS.lat.oct. 32 LEO DE BALNEOLIS (Levi ben Gersoß, 1288-1344): DE SINIRUS, CHORDIS ET ARCUBUS, Item instrumento revelatore secretorum, c. 1-9. In der Übersetzung des Petrus de Alexandria (1342). Am Ende des 4. Kapitels (15v-16r) eingeschoben: Additio seu scholia; endet: Vide ergo Leonem ludeum in tractatu suo de baculo Iacob. Ir: lnstrumentum quod reveLatorem secretorum nuncupant / Adi 22 Novembris Anno 1549 Lipsie ab Andrea Shonero / Simon Jacob Mathematicarum rerum Amator; Textschluß 29r: Explicit tractatus instrumenti Astronomie . .. Conclusi 22. Novembris Anno 1549 Lipsie. Zum Verfasser und zum Text s. G. SARTON Introduction to the history of science 3,1 (1947) 595, 598-602. Zu Andreas Schoner aus Nürnberg (1528-1590; 1548 immatr. in Leipzig) s. POGGENDORFF 2, 836; Die Matrikel der Universität Leipzig 1 (1894) 674; Diktator des Textes? - Simon Jacob aus Coburg (1510?-1564; NDB 10,219-220) starb als Rechenmeister in Frankfurt. Das autographe Manuskript seines "Rechenbuches", datiert Frankfurt 1. November 1552, und die H s. eines Visierruten-Traktats, datiert 1559, beide nach Ausweis von Setzerzeichen als Druckvorlagen benutzt, jetzt unter der Signatur: Ms. germ. oct. 79 (alt: IV 41) und Ms. germ. qu. 151. Ein mathematisches Handbuch aus Jacobs Besitz mit überwiegend eigenhändigen Aufzeichnungen aus den Jahren 1551-1561 gelangte ab haeredibus Simonis Jaeobi in die Bibliothek Johann Hartmann Beyers (Signa.tur: IV 47; jetzt: Ms. Ff. J. H. Beyer). 1549 . Leipzig . Schreiber und Besitzer: Sion Jacob (Schriftvergleich) . 1+ 31 BI. . 21 x 16 . auf den Blamälldern zahlreiche geometrische Zeichnungen . Buchblock in Langstichtechnik auf Pergamentmakulatur geheftet (Lat. asz.ctisd.er Text; 14. Jh.) . alte Signatur: IV, 45 . WALDSCHMIDT (um 1700) 207; LUCTUS (1728) Sect. XI S. 434; KRISTELLER Iter Italicum 3, 529; G. POWITZ pie datierten Handschriften der Stadt- und Universitätsbibliothek Frankfurt am Main (1984) 53 m. Abb. 307, 308 (Schreiber dort: Andreas Schollcr).
This paper describes the use of a Radial Basis Function (RBF) neural network in the approximation of process parameters for the extrusion of a rubber profile in tyre production. After introducing the rubber industry problem, the RBF network model and the RBF net learning algorithm are developed, which uses a growing number of RBF units to compensate the approximation error up to the desired error limit. Its performance is shown for simple analytic examples. Then the paper describes the modelling of the industrial problem. Simulations show good results, even when using only a few training samples. The paper is concluded by a discussion of possible systematic error influences, improvements and potential generalisation benefits. Keywords: Adaptive process control; Parameter estimation; RBF-nets; Rubber extrusion
Diese Arbeit plädiert für eine rationale Behandlung von Patientendaten und untersucht dazu die Analyse der Daten mit Hilfe neuronale Netze etwas näher. Erfolgreiche Beispielanwendungen zeigen, daß die menschlichen Diagnosefähigkeiten deutlich schlechter sind als neuronale Diagnosesysteme. Für das Beispiel der neueren Architektur mit RBF-Netzen wird die Funktionalität näher erläutert und gezeigt, wie menschliche und neuronale Expertise miteinander gekoppelt werden kann. Der Ausblick deutet Anwendungen und Praxisproblematik derartiger Systeme an.
The paper focuses on the division of the sensor field into subsets of sensor events and proposes the linear transformation with the smallest achievable error for reproduction: the transform coding approach using the principal component analysis (PCA). For the implementation of the PCA, this paper introduces a new symmetrical, lateral inhibited neural network model, proposes an objective function for it and deduces the corresponding learning rules. The necessary conditions for the learning rate and the inhibition parameter for balancing the crosscorrelations vs. the autocorrelations are computed. The simulation reveals that an increasing inhibition can speed up the convergence process in the beginning slightly. In the remaining paper, the application of the network in picture encoding is discussed. Here, the use of non-completely connected networks for the self-organized formation of templates in cellular neural networks is shown. It turns out that the self-organizing Kohonen map is just the non-linear, first order approximation of a general self-organizing scheme. Hereby, the classical transform picture coding is changed to a parallel, local model of linear transformation by locally changing sets of self-organized eigenvector projections with overlapping input receptive fields. This approach favors an effective, cheap implementation of sensor encoding directly on the sensor chip. Keywords: Transform coding, Principal component analysis, Lateral inhibited network, Cellular neural network, Kohonen map, Self-organized eigenvector jets.
After a short introduction into traditional image transform coding, multirate systems and multiscale signal coding the paper focuses on the subject of image encoding by a neural network. Taking also noise into account a network model is proposed which not only learns the optimal localized basis functions for the transform but also learns to implement a whitening filter by multi-resolution encoding. A simulation showing the multi-resolution capabilitys concludes the contribution.
We present a framework for the self-organized formation of high level learning by a statistical preprocessing of features. The paper focuses first on the formation of the features in the context of layers of feature processing units as a kind of resource-restricted associative multiresolution learning We clame that such an architecture must reach maturity by basic statistical proportions, optimizing the information processing capabilities of each layer. The final symbolic output is learned by pure association of features of different levels and kind of sensorial input. Finally, we also show that common error-correction learning for motor skills can be accomplished also by non-specific associative learning. Keywords: feedforward network layers, maximal information gain, restricted Hebbian learning, cellular neural nets, evolutionary associative learning
It is well known that artificial neural nets can be used as approximators of any continuous functions to any desired degree and therefore be used e.g. in high - speed, real-time process control. Nevertheless, for a given application and a given network architecture the non-trivial task remains to determine the necessary number of neurons and the necessary accuracy (number of bits) per weight for a satisfactory operation which are critical issues in VLSI and computer implementations of nontrivial tasks. In this paper the accuracy of the weights and the number of neurons are seen as general system parameters which determine the maximal approximation error by the absolute amount and the relative distribution of information contained in the network. We define as the error-bounded network descriptional complexity the minimal number of bits for a class of approximation networks which show a certain approximation error and achieve the conditions for this goal by the new principle of optimal information distribution. For two examples, a simple linear approximation of a non-linear, quadratic function and a non-linear approximation of the inverse kinematic transformation used in robot manipulator control, the principle of optimal information distribution gives the the optimal number of neurons and the resolutions of the variables, i.e. the minimal amount of storage for the neural net. Keywords: Kolmogorov complexity, e-Entropy, rate-distortion theory, approximation networks, information distribution, weight resolutions, Kohonen mapping, robot control.