Universitätspublikationen
Refine
Year of publication
- 2016 (178) (remove)
Document Type
- Doctoral Thesis (178) (remove)
Has Fulltext
- yes (178)
Is part of the Bibliography
- no (178)
Keywords
- Anerkennung (2)
- 19F NMR shifts for fluoroarenes (1)
- AF4 (1)
- ARDS (1)
- Activity (1)
- Acylimin Sulfonylimin (1)
- Adaptation (1)
- Aedes albopictus (1)
- Aktuelle Motivation (1)
- Allergie (1)
- Alzheimer (1)
- Amphisphaeriales (1)
- Angiogenesis (1)
- Arbeitsbedingungen (1)
- Asian Tiger Mosquito (1)
- Asiatische Tigermücke (1)
- Außerschulisches Lernen (1)
- Batten disease (1)
- Beatmung (1)
- Biodiversität (1)
- Biomoleküle (1)
- Blockplay (1)
- Broken symmetry (1)
- CCLM (1)
- CORDEX-EAS (1)
- Call Center (1)
- Carotinoidbiosynthese (1)
- Computerkompetenz (1)
- Conformational Dynamics (1)
- DDX6 (1)
- DFT (1)
- DPOAE (1)
- Delegation (1)
- Density Functional Theory (1)
- Deutschland (1)
- ESD (1)
- Early Childhood (1)
- Emittanz (1)
- Enzymes and Biomimetic Systems (1)
- Epigenetik (1)
- Error Mitigation (1)
- Erythropoetin (1)
- Exercise (1)
- Exercise Metabolism (1)
- Eye Tracking (1)
- FPGA (1)
- FRANZ-Projekt (1)
- Familie (1)
- Family (1)
- Fault Tolerance (1)
- Fettsäuresynthase (1)
- Forschungsbohrung Sichenhausen (1)
- Fragmentierungsquerschnitte (1)
- Freiheit (1)
- Gas Chromatography (1)
- Geoelektrik (1)
- Geometrie (1)
- Geometry (1)
- Germanistik (1)
- Geschichte 1966-2001 (1)
- Geschlechterforschung (1)
- Gespräch (1)
- Gesundheitliche Ungleichheit (1)
- Gesundheitsverhalten (1)
- Grundschule (1)
- HIV (1)
- Halocarbons (1)
- Halogenated Trace Gases (1)
- Handlung (1)
- Hausarztpraxis (1)
- Hegel (1)
- HepG2 Zellen (1)
- Hodgkin-Lymphom (1)
- Holzklötzchen (1)
- Hydroxylation (1)
- Immigration (1)
- Integral Geometry (1)
- Intelligence (1)
- Interaction (1)
- Invariant object recognition (1)
- Inversion (1)
- Ionenquelle (1)
- Ionizing Radiation (1)
- Iran (1)
- Juvenile neuronal ceroid lipofuscinosis (1)
- Karotte (1)
- Kinzigtal (1)
- Konstruktvalidität (1)
- Kurdistan (1)
- LINE-1 (1)
- LU-102 (1)
- Lernen (1)
- Lungenkontusion (1)
- Lösungsprozess (1)
- Material Flow Analysis (1)
- Mathematics Learning (1)
- Mathematik (1)
- Medizinische Fachangestellte (1)
- Mehrsprachigkeit (1)
- Mitochondria (1)
- Mitochondrium (1)
- Modellorganismus (1)
- Moesia Superior (1)
- Molarenevolution (1)
- Monsun (1)
- Multikomponentenreaktion (1)
- Mycophenolat Mofetil (1)
- NMR shift calculation (1)
- Nahrungsmittelallergie (1)
- Namibia (1)
- Nok Culture (1)
- Oligonukleotide (1)
- Optical Spectroscopy (1)
- Ostasien (1)
- P-TEFb (1)
- Palladium-Katalyse (1)
- Participation (1)
- Partizipation (1)
- Pb-Ag (1)
- Pepperpot (1)
- Pestalotia (1)
- Pestalotiopsis (1)
- Phaeodactylum tricornutum (1)
- Pharmacology (1)
- Photodynamics (1)
- Primärversorgung (1)
- Proteine (1)
- Prozedurales Wissen (1)
- Qualifikation (1)
- RCM (1)
- RNAPII (1)
- Rainwater Harvesting (1)
- Reasoning (1)
- Republik Moldau (1)
- Retrotransposition (1)
- Ribonucleic Acids (1)
- Roman (1)
- Rotorest (1)
- SORN (1)
- STARS (1)
- STDP (1)
- Saccharomyces cerevisiae (1)
- Schlitz-Gitter (1)
- Scientific Materials Analysis (1)
- Sellerie (1)
- Single Event Effects (1)
- Single-particle (1)
- Skeletal muscle fiber-type switching (1)
- Soziolinguistik (1)
- Spiel (1)
- Sportunterricht (1)
- Sprachliche Minderheiten (1)
- Staatstheorie (1)
- Stahlenschäden (1)
- Strafrecht (1)
- Strukturwandel (1)
- Support (1)
- Sustainability Evaluation (1)
- Talk (1)
- Technologiebasiertes Problemlösen (1)
- Terracotta Figurines (1)
- Time-of-Flight Mass Spectrometry (1)
- Time-resolved Fluorescence (1)
- Trace Gas Analysis (1)
- Transfer learning (1)
- Transition Metals (1)
- Turkish immigrants (1)
- Türkisch (1)
- Universa feature extraxtion (1)
- Unternehmensstrafrecht (1)
- Unterstützung (1)
- Valuation Theory (1)
- Verankerung (1)
- Verbandsstrafrecht (1)
- Vogelsberg (1)
- Water Reuse (1)
- Wirtschaftsstrafrecht (1)
- Working Memory (1)
- Zelle (1)
- Zellfusion (1)
- Zurechnung (1)
- Zweiphotonen (1)
- archaeometallurgy (1)
- categorisation-elaboration model (1)
- cell biology (1)
- cryo-EM (1)
- drug discovery (1)
- enantioselektive Synthese (1)
- episodic team process model (1)
- frühkindliche Erziehung (1)
- fungi (1)
- global and virtual teams (1)
- high-content screening (1)
- ibrutinib (1)
- ionisierende Strahlung (1)
- kinetische Therapie (1)
- lattice (1)
- magnetic exchange coupling constants (1)
- magnon condensation (1)
- magnon-phonon interactions (1)
- medieval/early modern (1)
- microRNA (1)
- molecular phylogenetics (1)
- multiple myeloma (1)
- neural networks (1)
- phase transitions (1)
- photolabile Schutzgruppe (1)
- plasticity (1)
- proteasom inhibitor (1)
- quantum chromodynamics (1)
- regionale Klimasimulation (1)
- steroid (1)
- synaptic normalization (1)
- team innovation (1)
- unterrichtliche Vor- und Nachbereitung (1)
- variability (1)
- yttrium-iron garnet (1)
- Übergewicht und Adipositas im Kindesalter (1)
Institute
- Biowissenschaften (38)
- Biochemie und Chemie (25)
- Physik (23)
- Medizin (17)
- Geowissenschaften (14)
- Informatik (11)
- Mathematik (9)
- Pharmazie (8)
- Biochemie, Chemie und Pharmazie (5)
- Kulturwissenschaften (5)
Eine möglichst realistische Abschätzung von Strahlenschäden ist von entscheidender Bedeutung im Strahlenschutz und für die Strahlentherapie. Die primären Strahlenschäden an der DNS werden heute mit Monte-Carlo-Codes berechnet. Diese Codes benötigen möglichst genaue Fragmentierungsquerschnitte verschiedenster biomolekularer Systeme als Eingangsparameter. Im Rahmen der vorliegenden Arbeit wurde ein Experiment aufgebaut, welches die Bestimmung der Fragmentierungsquerschnitte von Biomolekülen ermöglicht. Die einzelnen Baugruppen des Aufbaus wurden vor dem Beginn des Experimentes bezüglich ihrer Eigenschaften, die die Genauigkeit der Messergebnisse beeinflussen können, charakterisiert. Die Resultate dieser Experimente werden als Eingangsdaten für die Berechnung von primären strahleninduzierten Schäden in der DNS mit Hilfe von Monte-Carlo-Codes eingesetzt.
Eine besondere Herausforderung stellte die Präparation eines Überschallgasstrahls für biomolekulare Substanzen dar. Für die Präparation müssen die Targetsubstanzen zunächst in die Gasphase überführt werden. Im Falle von Biomolekülen ist diese Überführung auf Grund ihrer niedrigen Dampfdrücke bei Raumtemperatur und chemischen Reaktivität mit technischen Problemen verbunden. Die Probleme wurden mittels einer speziellen Konstruktion der Präparationseinrichtung, welche eine direkte Einleitung der Probensubstanzen in die vom Trägergas durchströmte Mischkammer ermöglicht, gelöst. Für die Genauigkeit der gemessenen Fragmentierungsquerschnitte spielen mehrere Faktoren eine Rolle. Neben dem Bewegungsprofil des Überschallgasstrahls, den kinetischen Energien der Fragmentionen und den ionenoptischen Eigenschaften des Flugzeitspektrometers beeinflusst die geometrische Beschaffenheit der Detektionszone maßgeblich die Genauigkeit des Experimentes. Die Position und Ausdehnung des sichtbaren Volumens sind nicht nur durch den Überlappungsbereich zwischen dem Elektronen- und dem Überschallgasstrahl bestimmt, sondern hängen auch von der kinetischen Energie der Fragmente ab. Für dessen Ermittlung wurden daher auch die Trajektorien der Fragmente simuliert. Bei den Experimenten an der PTB-Apparatur ist die frei wählbare Zeitdifferenz zwischen dem Auslösen eines Elektronenpulses und dem Absaugen der Fragmentionen ein wichtiger Messparameter. Ihr Einfluss auf die Messergebnisse wurde ebenfalls neben der Nachweiswahrscheinlichkeit des verwendeten Ionendetektors untersucht. Die Kalibrierung der Flugzeitspektren, d. h. die Umwandlung der Flugzeitspektren in Massenspektren erfolgte anhand der bekannten Flugzeitspektren von Edelgasen und Wasserstoff.
Nach der Charakterisierung der Einflussfaktoren und Kalibrierung der Flugzeitspektren wurden die energieabhängigen Fragmentierungsquerschnitte für Elektronenstoß von mehreren organischen Molekülen, darunter die von Modellmolekülen für die DNS-Bausteine gemessen. Die Flugzeitspektren von THF wurden mit der PTB-Apparatur für einige kinetische Energien der Elektronen in Abhängigkeit von der Zeitdifferenz zwischen dem Auslösen des Elektronenpulses und dem Starten der Analyse durchgeführt. Messungen von Pyrimidin wurden sowohl an der PTB-Apparatur als auch mit COLTRIMS durchgeführt. Die mit COLTRIMS gewonnenen Ergebnisse liefern wichtige Zusatzinformationen über die Fragmentierungsprozesse. COLTRIMS ermöglicht die Messung der zeitlichen Korrelationen zwischen den auftretenden Fragmentionen und damit tiefere Einblicke in die bei der Entstehung der Fragmente beteiligten Reaktionskanäle. Der Vorteil der PTB-Apparatur besteht darin, dass die relativen Auftrittswahrscheinlichkeiten aller Fragmentionen genauer bestimmt werden können.
From an early understanding of organisational theorist (Bartlett & Ghosal, 1989; 1990), the function of global teams in transnational organisations has been conceptualised as the transformation of different embedded cultural practices for the development of a global strategy, products and services. Simultaneously, in the field, from the beginning of the 1990ies to the edge of the new millennium neo-liberal political developments enforced a free flow of capital on a global level (cf. Turner, 2006). In line with the development of respective connectivity via the internet the form of globally distributed team work was spread (Maznevski & Chudoba, 2000). In a study by Biggs (2000), published just after the millennial change, it was shown that 60% of tasks in multinational companies are accomplished by virtual teams. More recent data (Society for Human Resource Management, 2012) showed that the prevalence of such teams stayed more or less constant in the last 10 years. According to the survey 66% of companies are working with distributed global teams.
Globally distributed teams were already described by Bartlett and Ghosal (1989) in their functions of articulation and translation of differing market practices for the integration of requirements and needs on a global level. From a European perspective the importance to further develop innovation capabilities in order to compete in the global market is stressed today (Imp3rove, 2012). In a globalised economy not only the big multinational companies are involved in globally distributed research and development activities (R&D). On the level SMEs, for example, in Switzerland the involvement in global development processes is increasing (Gassmann, 2009). From my own experiences in working with Swiss SMEs, the macro-economic processes in regard to the strong Swiss Franc may accelerate such processes. Thus, the form of globally distributed teams, and their functional task in global development processes, can be viewed as highly relevant, in a globalised economy.
The crucial question for companies at the moment is, if teams can be enabled for innovative project work, which enables the integration of diverging perspectives in a globally distributed setting? Or, if such teams have to be collocated for more innovative, interdependent task work? Requirements for integrating embedded knowledge from different regionally defined clusters into global innovations at least, seems to indicate for the relevance of interdependent globally distributed team work (cf. Li, Eden, Hitt, Ireland, & Garrett, 2012). Bilateral practices of partnering, for example in the Swiss pharmaceutical sector, lead to the integration of selected subsidiaries in the R&D process of the company (Festel et al., 2010). Thus, the form of dispersion for project teams becomes more critical for effective global R&D practices (Boyer O’Leary & Cummings, 2007). So called partially distributed teams integrating balanced subgroups between two sites, hence, become an important subject of inquiry with practical relevance.
The context of partially distributed team work represents by virtue a context involving multiple perspectives influenced by the involvement of actors stemming from different cultural contexts (Dekker, Rutte, & Van den Berg, 2008). It thereby provides the synergetic potential for integrating different perspectives in the resolution of complex problems on a global level (Janssens & Brett, 2006). Simultaneously, cultural diversity engenders challenges for collaboration. Challenges, like different understandings and interpretations regarding tasks, the structuring of communication (Maznevski & Chudoba, 2000) and unexpected events occurring in the collaboration between the actors (Dekker et al., 2008; Oertel & Antoni, 2014) were identified in respective empirical studies.
Opportunities and challenges of partially distributed global teams can be compared with the problematic of face-to-face (f2f) teams with a moderate amount of diversity. Studies have shown (see Thatcher & Patel, 2011 for a meta-analysis) that when the distribution of diversity characteristics is aligned to potentially form culture specific subgroups, so-called diversity faultlines (Lau & Murnighan, 1998), negative subgroup dynamics are reinforced.
To achieve the above mentioned synergetic potentials it seems important to effectively cope with such negative dynamics and allow for a balanced participation in partially distributed teams (Janssens & Brett, 2006). In the research on faultline teams, especially the structuring of task-related interdependences across respective subgroups has been identified as an important impediment for the mentioned subgroup dynamics. Task interdependences, which cross functional roles across respective group faultlines (Bettencourt, Molix, Talley, & Eubanks, 2007; Marcus-Newhall, Miller, Holtz, & Brewer, 1993), are able to unlock the inherent potentials of globally distributed teams on more complex tasks that require the integration of different perspectives. From a work group diversity perspective (van Knippenberg & Schippers, 2007), partially distributed global teams represent a research object for studying the interaction between social categorisation processes involved in the above mentioned subgroup dynamics, and processes of task-related information processing required for innovative team outputs. The exploration of effects of task structures on the interaction between categorisation processes and task-related information processing (van Knippenberg et al., 2004), will be in the main explorative research focus of this thesis. The research thesis represents a heuristic explorative inquiry (Kleining & Witt, 2001) of respective dynamics and structural as well as process-related enablers.
The thesis starts with the theoretical part, in which the historical development of the understanding of teams as open, complex and temporally dynamic systems (Arrow et al., 2005, 2000), will be outlined. A sound definition of partially distributed global teams, including the respective contextual characteristics will be delineated. In a sensitizing framework (Blumer, 1954) which guided the explorative research process, the central boundary condition of task interdependence (Wageman, 2001) and respective episodic theories for explaining global task-related dynamics in teams (Marks et al., 2001), the dynamics of social categorisation (Gaertner, Dovidio, Anastasio, Bachman, & Rust, 1993; Gaertner & Dovidio, 2000), as well as the interaction between social categorisation processes and task-related information processing will be integrated (van Knippenberg et al., 2004). According to the framework, empirical studies on effects of task interdependence on interactions between task-related information processing and social categorisation processes will be addressed (van Knippenberg et al., 2004).
The empirical part of the contribution will be split in two parts. In the first heuristic exploratory study eleven partially distributed global teams are followed up during the time of relevant innovation projects. The approach allowed the study of task interdependence, productive interactions with social categorisation processes and there effects on team innovation. In the second empirical step, the developed hypotheses, were tested in an experimental simulation (Arrow et al., 2005, 2000) in undergraduate courses.
As a conclusion of the two exploratory studies, an episodic team process model will be outlined. The model specifies interdependence dynamics, which allows for team innovation. Furthermore, on a process level, the episodic categorisation-elaboration model (van Knippenberg et al., 2004) proposes three critical team performance episodes. Dynamics in the interplay between task-related information processing and social categorisation processes allow for the development of hypothesis for further research projects. Finally the implications for theory and the practical relevance of the heuristic model will be discussed.
In Belize, which is well known for the Belize Barrier Reef and its offshore atolls, coastal lagoons are frequent morphological features along the coast. They represent transitional environments between siliciclastic and carbonate settings. In order to shed light into the Holocene evolution of coastal lagoon environments, five localities along the central coast of Belize were selected as coring sites. These include four coastal lagoons and one marsh area, namely Mantatee Lagoon, Mullins River Beach, Colson Point Lagoon, Commerce Bight Lagoon and Sapodilla Lagoon. A total of 26 sediment cores with core lengths ranging from 109 cm to 500 cm, were drilled using a Lanesky-vibracorer. Overall, 73 m of Holocene sediments and Pleistocene soil were recovered. Together with 58 radiocarbon dates the sediments reveal details on the sediment architecture and depositional features of the localities.
Marine inundation of the mainland and coastal lagoon formation started around 6 kyrs cal BP.
As a response to sea-level rise during the Holocene transgression, facies retrograded towards the coast, as seen in marginal marine overlying brackish mollusk faunas. Evidence for late Holocene progradation of facies due to sea-level stagnation is largely lacking. The occurrence of landward thinning sand beds, hiatuses and marine fauna in lagoonal successions are indications of event (overwash) sedimentation. Sediments recovered are largely of Holocene age (<7980 cal BP), overlying Pleistocene sections. Analyses of sediment composition and texture, radiocarbon dating and mollusk shell identification were used to describe and correlate sedimentary facies.
XRD analyses have identified quartz as the dominant mineral, with the Maya Mountains as main source of coastal lagoon sediments. The most common sedimentary facies include peat and peaty sediment, mud, sand, and poorly sorted sediments. Pleistocene soil forms the basement of Holocene sediments. Holocene mud represents lagoon background permanent sedimentation.
Peats and peat-rich sequences were deposited in mangrove swamp environments, whereas sandy facies mainly occur in the shoreface, beach, barriers, bars, barrier spits and overwash deposits. Facies successions could be identified for each locality, but it has proven difficult to correlate the stratigraphic sequences, especially among localities. These differences among the five locations studied suggest that apart from regional influence such as sea-level rise, local environmental factors such as small-scale variation in geomorphology and resulting facies heterogeneity, connectivity of the lagoon with the sea, antecedent topography and river discharge, were responsible for coastal sedimentation and lagoon development in the Holocene of Belize.
Faunal composition and distribution patterns of mollusk assemblages from 20 shell concentrations in cores collected in coastal lagoons, a mangrove-fringed tidal inlet and the marginal marine area (shallow subtidal) show considerable variation due to environmental heterogeneity and the interplay of several environmental factors in the course of the mid-late Holocene (ca. 6000 cal BP to modern). The investigated fauna ≥2 mm comprises 2246 bivalve, gastropod and 11 scaphopod specimens. Fifty-three mollusk species, belonging to 42 families, were identified. The bivalve Anomalocardia cuneimeris and cerithid gastropods are the dominant species and account for 78% of the total fauna. Diversity indices are low in concentrations from lagoons and relatively high in the marginal marine and tidal inlet areas.
Based on cluster analysis and nonmetric multidimensional scaling (NMDS), seven lagoonal assemblages and three marginal marine/tidal inlet assemblages were defined. A separation between lagoonal and marginal marine/tidal inlet assemblages seen in ordination indicates a lagoon-onshore gradient. The statistical separation among lagoonal assemblages demonstrates environmental changes during the Holocene evolution of the coastal lagoons, which is probably related to the formation of barriers and spits. The controlling factors of species distribution patterns are difficult to figure out, probably due to the heterogeneity of the barrier-lagoon systems and the interaction of paleoecological and paleoenvironmental factors. In addition to the taxonomic analysis, a taphonomic analysis of 1827 valves of A. cuneimeris from coastal lagoons was carried out. There is no relationship between depth and age of shells and their taphonomic condition. Size-frequency distributions and right-left valve ratios of A. cuneimeris suggest that valves were not transported over long distances but were deposited parautochthonously in their original habitat. Shells from tidal inlet and marginal marine environments were also predominantly deposited in their original habitats.
Since the Belize coast was repeatedly affected by hurricanes and the paleohurricane record for this region is poor, the sediment cores have been examined in order to identify storm deposits.
The paleohurricane record presented in this study spans the past 8000 years and exhibits three periods with increased evidences of hurricane strikes occurring at 6000-4900 cal yr BP, 4200-3600 cal yr BP and 2200-1500 cal yr BP. Two earlier events around 7100 and 7900 cal yr BP and more recent events around 180 cal yr BP and during modern times have been detected. Sand layers, redeposited corals and lagoon shell concentrations have been used as proxies for storm deposition. Additionally, hiatuses and reversed ages may indicate storm influence. While sand layers and corals represent overwash deposits, the lagoon shell concentrations, which mainly comprise the bivalve Anomalocardia cuneimeris and cerithid gastropods, have been deposited due to changes in lagoon salinity during and after storm landfalls. Comparison with other studies reveals similarities with one record from Belize, but hardly any matches with other published records. The potential for paleotempestology reconstructions of the barrier-lagoon complexes along the central Belize coast differs depending on geomorphology, and deposition of washovers in the lagoon basins is limited, probably due to the interplay of biological, geological and geomorphological processes.
Water is scarce in semi-arid and arid regions. Using alternative water sources (i.e. non-conventional water sources), such as municipal reuse water and harvested rain, contributes to using existing water resources more efficiently and productively. The aim of this study is to evaluate the two alternative water sources reuse water and harvested rain for the irrigation of small-holder agriculture from a system perspective. This helps decision and policy makers to have proper information about which system and technology to adopt under local conditions. For this, the evaluation included ecologic, societal, economic, institutional and political as well as technical aspects. For the evaluation, the study area in central-northern Namibia was chosen in the frame of the research and development project CuveWaters. The main methods used include a mathematical material flow analysis, the computation and modelling of crop requirements, a multi-criteria decision analysis using the Analytical Hierarchy Process (AHP) method and a financial cost-benefit analysis. From a systemic perspective, the proposed novel systems were compared to the exciting conventional infrastructure. The results showed that both water reuse and rainwater harvesting systems for the irrigation of small-holder horticulture offer numerous technological, ecologic, economic, societal, institutional and political benefits. Rainwater harvesting based gardens have a positive benefit-cost ratio under favorable conditions. Government programs could fund the infrastructure investment costs, while the micro-entrepreneur can assume a micro-credit to finance operation and maintenance costs. Installing sanitation in informal settlements and reusing municipal water for irrigation reduces the overall water demand of households and agriculture by 39%, compared to improving sanitation facilities in informal settlements without reusing the water for agriculture. Given that water is the limiting factor for crop fertigation, the generated nutrient-rich reuse water is sufficient to annually irrigate about 10 m2 to 13 m2 per sanitation user. Compared to crop nutrient requirements, there are too many nutrients in the reuse water. Thus when using nutrient-rich reuse water, no use of fertilizers and a careful salt management is necessary. When comparing this novel system with improved sanitation, advanced wastewater treatment and nutrient-rich water reuse to the conventional and to two adapted systems, results showed that the novel CuveWaters system is the best option for the given context in a semi-arid developing country. Therefore, the results of this study suggest a further roll-out of the novel CuveWaters system. The methodology developed and the results of this study demonstrated that taking sanitation users into consideration plays a major role for the planning of an integrated water reuse infrastructure because they are the determinant factor for the amount of available nutrient-rich reuse water. In addition, it could be shown that water reuse and rainwater harvesting systems for the irrigation of small-scale gardens provide a wide range of benefits and can be key to using scarce water resources more efficiently and to contributing to the Sustainable Development Goals.
Im Fokus der vorliegenden Arbeit standen die Landschaftsrekonstruktion und die Abschätzung der Mensch-Umwelt Wechselwirkungen im Umfeld der kupferzeitlichen (ca. 5. Jts. v. Chr.) Siedlung Magura Gorgana im heutigen Süd-Rumänien. Im Zuge der sedimentologischen Untersuchungen wurde deutlich, dass zur Zeit der kupferzeitlichen Besiedlung in der Donauaue ein Paläosee existierte, welcher sich nahezu über den gesamten Auenbereich im Untersuchungsgebiet erstreckte. Mit der Entdeckung des Paläosees ‚Lacul Gorgana‘ ergaben sich, durch die exzellenten Eigenschaften von Seesedimenten als Geoarchiv, neue Möglichkeiten zur Paläomilieurekonstruktion und zur Abschätzung von Mensch-Umwelt-Interaktionen. Bis etwa in das 8. Jt. v. Chr. war die Donau in diesem Gebiet durch ein fluviales System charakterisiert, welches vermutlich aus einer Vielzahl von Gerinnen bestand und einem ‚braided river‘ ähnelte. Während des 8. Jt. v. Chr. begann die Bildung des Paläosees ‚Lacul Gorgana‘. Die Gründe hierfür sind unbekannt, wenngleich ein Zusammenhang zum Anstieg des Schwarzmeerspiegels in diesem Zeitraum naheliegt. Spätestens ab dem 13. Jh. n. Chr. kommt es zur Aggradation- bzw. Verlandung des Paläosees ‚Lacul Gorgana‘. Infolge progradierender Zuflüsse wurde die einstige Seefläche in kleinere Seen fragmentiert. Dies dauerte bis zu den Trockenlegungsmaßnahmen in den 1960er Jahren an. Somit ist die Verlandung zum gegenwärtigen Zeitpunkt, zumindest theoretisch, noch nicht abgeschlossen. Die Ausprägung des rezenten Flussbettes der Donau begann spätestens mit der Verlandung des Paläosees, genaue chronologische Angaben sind jedoch anhand der aktuellen Datenlage nicht möglich. Auf der Grundlage geochemischer Untersuchungen geben die Sedimente des Paläosees ‚Lacul Gorgana‘ Hinweise auf alternierende Bedingungen bezüglich aerober und anaerober Akkumulationsmilieus. Dabei sind die aeroben Abschnitte durch einen höheren Anteil klastischen Materials, einem niedrigeren Anteil organischen Materials und allgemein helleren Sedimenten gegenüber den anaeroben, allgemein dunkleren Abschnitten gekennzeichnet. Die Gesamtheit der sedimentologischen Befunde, und der Vergleich mit Untersuchungen anderer Autoren in benachbarten Einzugsgebieten legen nahe, dass das alternierende Seemilieu in erster Linie durch Variationen der klimatischen Bedingungen im Einzugsgebiet der Donau verursacht wurde. Diese Variationen führten zur Veränderung der Stärke von Erosion und dem Charakter des erodierten Materials. Der hohe, zeitlich begrenzte Eintrag organisch gebundenen Phosphates in Bereichen der unteren Dunklen Lage, welcher weitestgehend zeitgleich zur neolithischen und kupferzeitlichen Besiedlung akkumuliert wurde und mit hoher Wahrscheinlichkeit zu einer Eutrophierung des ‚Lacul Gorganas‘ führte, lässt sich vermutlich durch die menschliche Besiedlung der Uferzone des Sees in diesem Zeitraum erklären. Im Verlauf der Untersuchungen im Einzugsgebiet erwies sich der ‚Regionalisierungsansatz‘ als effektives und unabdingbares Werkzeug bei der Differenzierung der Seegenese. Diese Methode dient der Relativierung bzw. Abschätzung der raumzeitlichen Heterogenität des Akkumulationsmilieus anhand des Vergleiches von Sedimentstratigraphien bzw. Sedimentbohrkernen von unterschiedlichen Standorten innerhalb des Paläosees. So konnten die alternierenden Milieubedingungen innerhalb der Seesedimente deutlich als allgemeine und nicht nur als lokale Eigenschaft einzelner Bereiche des Paläosees Lacul Gorgana identifiziert werden. Daneben zeigte die Regionalisierung, dass die aeroben Bereiche nicht nur durch einen erhöhten klastischen Eintrag gekennzeichnet sind, sondern dass dieser auch in Richtung des nördlichen Litorals zunimmt. Dies spricht für eine Herkunft des Materials aus dieser Richtung und erlaubt eine räumlich differenzierte Betrachtung der Seegenese. Darüber hinaus ermöglicht der ‚Regionalisierungsansatz‘, die räumliche Variabilität bestimmter Parameter, beispielsweise des TOC/TN Anstiegs in Richtung Litoral, zu relativieren. Diese Relativierung trägt zum besseren Verständnis spezifischer Prozessabläufe bei. Während der Untersuchung wurde ebenfalls deutlich, dass das Geoarchiv Seesediment eine Vielzahl verschiedenster Signale unterschiedlichster Genese aus dem Einzugsgebiet und dem Gewässermilieu selbst als Proxy-Information speichert. Die Überlagerung dieser Signale innerhalb der Sedimentstratigraphie erschwert mitunter die Identifikation einzelner Prozesse oder Prozesskaskaden. In diesem Zusammenhang erweist sich der ‚Regionalisierungsansatz‘ erneut als sinnvolles Hilfsmittel, da über diesen eine Signaldifferenzierung erfolgen kann, unter der Annahme, dass es eher unwahrscheinlich ist, dass alle Signale in allen Bereichen des Milieus in gleicher Intensität vorliegen. Gerade für die Untersuchung sowohl allochthoner als auch autochthoner Ereignissedimentation ist jedoch die Differenzierung zur ‚Hintergrundsedimentation‘ unabdingbar für ein umfassendes Prozessverständnis.
This thesis contributes to the field of machine learning with a specific focus on the methods for learning relations between the inputs. Learning relationships between images is the most common primitive in vision. There are many vision tasks in which relationships across images play an important role. Some of them are motion estimation, activity recognition, stereo vision, multi-view geometry and visual odometry. Many of such tasks mainly depend on motion and disparity cues, which are inferred based on the relations across multiple image pairs. The approaches presented in this thesis mainly deal with, but are not limited to, learning of the representations for motion and depth. This thesis by articles consists of five articles which present relational feature learning models along with their applications in computer vision. In the first article, we present an approach for encoding motion in videos. To this end, we show that the detection of spatial transformations can be viewed as detection of coincidence or synchrony between the given sequence of frames and a sequence of features which are related by the transformation we wish to detect. Learning to detect synchrony is possible by introducing "multiplicative interactions'' into the hidden units of single layered sparse coding models.
We show that the learned motion representations employed for the task of activity recognition achieve competitive performance on multiple benchmarks. Stereo vision is an important challenge in computer vision and useful for many applications in that field. In the second article, we extend the energy based learning models, which were previously used for motion encoding, to the context of depth perception. Given the common architecture of the models for encoding motion and depth, we show that it is possible to define a single model for learning a unified representation for both the cues. Our experimental results show that learning a combined representation for depth and motion makes it possible to achieve state-of-the-art performance at the task of 3-D activity analysis, and to perform better than the existing hand-engineered 3-D motion features. Autoencoder is a popular unsupervised learning method for learning efficient encoding for a given set of data samples. Typically, regularized autoencoders which are used to learn over-complete and sparse representations for the input data, were shown to fail on intrinsically high dimensional data like videos. In the third article, we investigate the reason for such a behavior. It can be observed that the regularized autoencoders typically learn negative hidden unit biases. We show that the learning of negative biases is the result of hidden units being responsible for both the sparsity and the representation of the input data. It is shown that, as a result, the behavior of the model resembles clustering methods which would require exponentially large number of features to model intrinsically high dimensional data. Based on this understanding, we propose a new activation function which decouples the roles of hidden layer and uses linear encoding. This allows to learn representations on data with very high intrinsic dimensionality. We also show that gating connections in the bi-linear models and the single layer models from articles one and two of this thesis can be thought of as a way to attain a linear encoding scheme which allows them to learn good representations on videos. Visual odometry is the task of inferring egomotion of a moving object from visual information such as images and videos. It can primarily be used for the task of localization and has many applications in the fields of robotics and navigation. The work in article four was motivated by the idea of using deep learning techniques, which are successful methods for many vision tasks, for visual odometry. The visual odometry task mainly requires inference of motion and depth information from visual input which can then be mapped to velocity and change in direction. We use relational feature models presented in the articles one and two for inferring a combined motion and depth representation from stereo video sequences. The combined representation is then mapped to discrete velocity and change in direction labels using convolutional neural networks. Our approach is an end-to-end deep learning-based architecture which uses a single type of computational model and learning rule. Preliminary results show that the architecture is capable of learning the mapping from input video to egomotion. Activity recognition is a challenging computer vision task with many real world applications. It is well know that it is a hard task to use computer vision research for real-time applications. In the fifth article of this thesis, we present a real-time activity recognition system based on deep learning based methods. Our approach uses energy based relational feature learning models for the computation of local motion features directly from videos. A bag-of-words over the local motion features is used for the analysis of activity in a given video sequence. We implement this system on a distributed computational platform and demonstrate its performance on the iCub robot. Using GPUs we demonstrate real time performance which makes the deployment of activity recognition systems in real world scenarios possible.
The phenomenon of magnetism has been known to humankind for at least over 2500 years and many useful applications of magnetism have been developed since then, starting from the compass to modern information storage and processing devices. While technological applications are an important part of the continuing interest in magnetic materials, their fundamental properties are still being studied, leading to new physical insights at the forefront of physics. The magnetism of magnetic materials is a pure quantum effect due to the electrons that carry an intrinsic spin of 1/2. The physics of interacting quantum spins in magnetic insulators is the main subject of this thesis.We focus here on a theoretical description of the antiferromagnetic insulator Cs2CuCl4. This material is highly interesting because it is a nearly ideal realization of the two-dimensional antiferromagnetic spin-1/2 Heisenberg model on an anisotropic triangular lattice, where the Cu(2+) ions carry a spin of 1/2 and the spins interact via exchange couplings. Due to the geometric frustration of the triangular lattice, there exists a spin-liquid phase with fractional excitations (spinons) at finite temperatures in Cs2CuCl4. This spin-liquid phase is characterized by strong short-range spin correlations without long-range order. From an experimental point of view, Cs2CuCl4 is also very interesting because the exchange couplings are relatively weak leading to a saturation field of only B_c=8.5 T. All relevant parts of the phase diagram are therefore experimentally accessible. A recurring theme in this thesis will be the use of bosonic or fermionic representations of the spin operators which each offer in different situations suitable starting points for an approximate treatment of the spin interactions. The methods which we develop in this thesis are not restricted to Cs2CuCl4 but can also be applied to other materials that can be described by the spin-1/2 Heisenberg model on a triangular lattice; one important example is the material class Cs2Cu(Cl{4-x}Br{x}) where chlorine is partially substituted by bromine which changes the strength of the exchange couplings and the degree of frustration.
Our first topic is the finite-temperature spin-liquid phase in Cs2CuCl4. We study this regime by using a Majorana fermion representation of the spin-1/2 operators motivated by theoretical and experimental evidence for fermionic excitations in this spin-liquid phase. Within a mean-field theory for the Majorana fermions, we determine the magnetic field dependence of the critical temperature for the crossover from spin-liquid to paramagnetic behavior and we calculate the specific heat and magnetic susceptibility in zero magnetic field. We find that the Majorana fermions can only propagate in one dimension along the direction of the strongest exchange coupling; this reduction of the effective dimensionality of excitations is known as dimensional reduction.
The second topic is the behavior of ultrasound propagation and attenuation in the spin-liquid phase of Cs2CuCl4, where we consider longitudinal sound waves along the direction of the strongest exchange coupling. Due to the dimensional reduction of the excitations in the spin-liquid phase, we expect that we can describe the ultrasound physics by a one-dimensional Heisenberg model coupled to the lattice degrees of freedom via the exchange-striction mechanism. For this one-dimensional problem we use the Jordan-Wigner transformation to map the spin-1/2 operators to spinless fermions. We treat the fermions within the self-consistent Hartree-Fock approximation and we calculate the change of the sound velocity and attenuation as a function of magnetic field using a perturbative expansion in the spin-phonon couplings. We compare our theoretical results with experimental data from ultrasound experiments, where we find good agreement between theory and experiment.
Our final topic is the behavior of Cs2CuCl4 in high magnetic fields larger than the saturation field B_c=8.5 T. At zero temperature, Cs2CuCl4 is then fully magnetized and the ground state is therefore a ferromagnet where the excitations have an energy gap. The elementary excitations of this ferromagnetic state are spin-flips (magnons) which behave as hard-core bosons. At finite temperatures there will be thermally excited magnons that interact via the hard-core interaction and via additional exchange interactions. We describe the thermodynamic properties of Cs2CuCl4 at finite temperatures and calculate experimentally observable quantities, e.g., magnetic susceptibility and specific heat. Our approach is based on a mapping of the spin-1/2 operators to hard-core bosons, where we treat the hard-core interaction by the self-consistent ladder approximation and the exchange interactions by the self-consistent Hartree-Fock approximation. We find that our theoretical results for the specific heat are in good agreement with the available experimental data.