Reinforcement learning on explicitly specified time scales

In recent years hierarchical concepts of temporal abstraction have been integrated in the reinforcement learning framework to improve scalability. However, existing approaches are limited to domains where a decomposition into subtasks is known a priori. In this article we propose the concept of explicitly selecting time scale related abstract actions if no subgoal related abstract actions are available. This concept is realised with multi-step actions on different time scales that are combined in one single action set. We exploit the special structure of the action set in the MSA-Q-learning algorithm. This approach is suited for learning optimal policies in unstructured domains where a decomposition into subtasks is not known in advance or does not exist at all. By learning different explicitly specified time scales simultaneously, we achieve a considerable improvement of learning speed, which we demonstrate on several benchmark problems.     

DDT-related compounds bound to the nonextractable particulate matter in sediments of the Teltow Canal,Germany

Sediment samples of the Teltow Canal (Berlin, Germany) were analyzed with respect to extractable and nonextractable organic compounds. The study focused on the identification and quantitation of bound 2,2-bis(chlorophenyl)-1,1,1-trichlorethane (DDT) residues in order to obtain further information about the fate of DDT-derived compounds within the particulate matter of the aquatic environment. Various chemical degradation techniques and a complementary online pyrolysis-GC/MS method were applied to the pre-extracted sediment residues. Generally, the distribution of the bound DDT-related compounds was found to differ distinctly from the substances distribution within the extractable fraction. The main metabolite of the anaerobic degradation pathway (2,2-bis(chlorophenyl)-1,1-dichlorethane, DDD) is most abundant in the sediment extracts but occurred only in insignificant concentrations in the degradation products of all procedures applied. The most abundant DDT-metabolites released after the degradation procedures were 4,4'-DBP, 4,4'-DDA, and 4,4'-DDM. In addition, 4,4'-DDM was detected at rather high concentrations by pyrolytic analysis. The results imply a weak association to the nonextractable particulate matter based on noncovalent interactions for the observed DDT-related contaminants. The release of these compounds was initiated by the modification and degradation of the organic macromolecular matrix as well as of the inorganic material. Furthermore, numerous methoxychlor-related compounds were detected not only in the extracts but also in parts of the hydrolysis products.     

Interferometry at the physical limit: how to measure sub-parts-per-million optical homogeneity in fused silica

Inspection of the refractive-index distribution in fused silica is very sensitive to thermally induced measurement errors. A model is derived for the estimation and interpretation of thermal errors applicable to interferometric homogeneity investigations. The outlines of the model are supported by experimental investigations and numerical calculations. The results state a mandatory temperature stability of deltaT = 0.02 K for a required reproducibility of sigma(delta(n)) < or = 1 x 10(-7) and a lower sensitivity of higher-order Zernike terms. Requirements of the interferometer environment include spatial and temporal stability. Only a small part of the frequency spectrum of temporal instabilities contributes significantly to the measurement error and is therefore critical for the system. Experimental values are given for different environmental conditions.     

Ultrasensitive detection of unstained proteins in acrylamide gels by native UV fluorescence

Visualization of proteins inside acrylamide and other gels usually relies on different staining methods. To omit the protein-staining procedure, we visualized unstained proteins inside acrylamide gels by laser excitation with ultraviolet (UV) light (280 nm, 35 mJ/cm2) and directly detected native UV fluorescence. In one-dimensional gels, a detection limit as low as 1 ng for bovine serum albumin and 5 ng for other proteins with a linear dynamic range (2.7 orders of magnitude) comparable to state of the art fluorescent dyes could be achieved. In addition, the application of this method to 20 microg of a whole cell lysate separated in a two-dimensional gel showed more than 600 spots. Since protein labeling always represents a serious obstacle in protein identification technologies, the working efficiency with our procedure can be considered as a significant improvement for protein visualization and reproducibility in proteomics.     

Complex dewetting scenarios captured by thin-film models

In the course of miniaturization of electronic and microfluidic devices, reliable predictions of the stability of ultrathin films have a strategic role for design purposes. Consequently, efficient computational techniques that allow for a direct comparison with experiment become increasingly important. Here we demonstrate, for the first time, that the full complex spatial and temporal evolution of the rupture of ultrathin films can be modelled in quantitative agreement with experiment. We accomplish this by combining highly controlled experiments on different film-rupture patterns with computer simulations using novel numerical schemes for thin-film equations. For the quantitative comparison of the pattern evolution in both experiment and simulation we introduce a novel pattern analysis method based on Minkowski measures. Our results are fundamental for the development of efficient tools capable of describing essential aspects of thin-film flow in technical systems.     

Synthesis of cubic zirconium and hafnium nitride having Th3P4 structure

High-pressure synthesis is a powerful method for the preparation of novel materials with high elastic moduli and hardness. Additionally, such materials may exhibit interesting thermal, optoelectronic, semiconductuing, magnetic or superconducting properties. Here, we report on the high-pressure synthesis of zirconium and hafnium nitrides with the stoichiometry M3N4, where M = Zr, Hf. Synthesis experiments were performed in a laser-heated diamond anvil cell at pressures up to 18 GPa and temperatures up to 3,000 K. We observed formation of cubic Zr3N4 and Hf3N4 (c-M3N4) with a Th3P4-structure, where M-cations are eightfold coordinated by N anions. The c-M3N4 phases are the first binary nitrides with such a high coordination number. Both compounds exhibit high bulk moduli around 250 GPa, which indicates high hardness. Moreover, the new nitrides, c-Zr3N4 and c-Hf3N4, may be the first members of a larger group of transition metal and/or lanthanide nitrides with interesting ferromagnetic or superconducting behaviour.     

Robustness of radial basis functions

Neural networks are intended to be used in future nanoelectronic technology since these architectures seem to be robust to malfunctioning elements and noise in its inputs and parameters. In this work, the robustness of radial basis function networks is analyzed in order to operate in noisy and unreliable environment. Furthermore, upper bounds on the mean square error under noise contaminated parameters and inputs are determined if the network parameters are constrained. To achieve robuster neural network architectures fundamental methods are introduced to identify sensitive parameters and neurons.     

Does age make a difference? Predicting physical activity of South Koreans.

Social cognition models of health behavior are commonly understood as being universal, which implies that they are applicable to groups varying in age or cultural background, for example. Cultural uniqueness and characteristics of life-span development, however, necessitate the study of differential effects. Accordingly, the health action process approach (HAPA) was examined in younger and middle-aged/older adults from South Korea (N = 697) who participated in a longitudinal health screening study with a 6-month time lag. The HAPA model had a good fit within the middle-aged/older adult sample. Physical activity was predicted by planning, coping self-efficacy, and intention, which were, in turn, predicted by action self-efficacy, outcome expectancies, and risk perceptions. Conversely, the results indicated a poor model fit in the younger adult sample. The results suggest a different motivation for the involvement in physical activity as a function of age. (PsycINFO Database Record (c) 2010 APA, all rights reserved)     

Material forces for inelastic models at large strains: application to fracture mechanics

Elastomeric materials show a wide range of different elastic and inelastic properties. Additionally, this class of materials is subjected to large deformations. Considering all these effects, fracture mechanical investigations are very challenging tasks and cannot be performed with standard approaches. Effects of inhomogeneities and discontinuities such as cracks can be investigated with the so-called material force approach in an efficient and elegant way. For comprehensive investigations of inelastic materials, the complete balance of the material motion problem has to be formulated. In this case, the material volume forces depend on the internal history variables which are required for the inelastic constitutive model. This paper derives a general formulation for rate-dependent and rate-independent inelastic materials based on a multiplicative split of the deformation gradient to cover viscoelastic and elastoplastic materials at finite deformations.     

Monitoring post-tsunami vegetation recovery in Phang-Nga province,Thailand, based on IKONOS imagery and field investigations – a contribution to the analysis of tsunami vulnerability of coastal ecosystems

A major tsunami in December 2004 devastated the coastal ecosystems along the Andaman Sea coast of Thailand. Since intact coastal ecosystems provide many important services for local communities at the Andaman Sea, it is crucial to investigate to what extent (in terms of percentage area and speed) the affected ecosystems were capable of recovering after the tsunami. Field measurements and multi-date IKONOS imagery were used to estimate the recovery and succession patterns of coastal vegetation types in the Phang-Nga province of Thailand, three years after the tsunami. Thus, this study contributes to a holistic understanding of the ecological vulnerability of the coastal area to tsunamis. A zone-based change detection approach is applied by comparing two change detection techniques: the first method involves the calculation of a recovery rate based on multi-temporal TNDVI (transformed normalized difference vegetation index) images (TNDVI approach), whereas the second approach is a combined approach of the change vector analysis (CVA). Although these two methods provide different types of information (quantitative for the TNDVI approach, qualitative for the CVA), they are comparable in terms of results and accuracies. The results reveal that recovery processes vary based on the type of the ecosystem and, furthermore, are strongly influenced by human activities. Grasslands, coconut plantations and the mixed vegetation cover could recover faster than the mangroves and casuarina forests. Among the forest ecosystems, recovery rates of casuarina forests were higher than for mangroves, but the recovery area was smaller. This study also discusses the potential and some limitations and inaccuracies of applying high-resolution optical imagery for assessing vegetation recovery at a local scale.