Implementation of a Neural Network Classifier for Noise Sources in the Ocean

The paper investigates the feasibility of implementing an intelligent classifier for noise sources in the ocean, with the help of artificial neural networks, using higher order spectral features. Non-linear interactions between the component frequencies of the noise data can give rise to certain phase relations called Quadratic Phase Coupling (QPC), which cannot be characterized by power spectral analysis. However, bispectral analysis, which is a higher order estimation technique, can reveal the presence of such phase couplings and provide a measure to quantify such couplings. A feed forward neural network has been trained and validated with higher order spectral features

Die Rolle des Zelladhäsionsmoleküls L1 in der Tumorprogression

Das neuronale Adhäsionsmolekül L1 wird neben den Zellen des Nervensystems auf vielen humanen Tumoren exprimiert und ist dort mit einer schlechten Prognose für die betroffenen Patienten assoziiert. Zusätzlich zu seiner Funktion als Oberflächenmolekül kann L1 durch membranproximale Spaltung in eine lösliche Form überführt werden. In der vorliegenden Arbeit wurde der Einfluss von L1 auf die Motilität von Tumorzellen untersucht. Lösliches L1 aus Asziten führte zu einer Integrin-vermittelten Zellmigration auf EZM-Substraten. Derselbe Effekt wurde durch Überexpression von L1 in Tumorlinien beobachtet. Weiterhin führt die L1-Expression zu einer erhöhten Invasion, einem verstärkten Tumorwachstum in NOD/SCID Mäusen und zur konstitutiven Aktivierung der MAPK ERK1/2. Eine Mutation in der zytoplasmatischen Domäne von hL1 (Thr1247Ala/Ser1248Ala)(hL1mut) führte hingegen zu einer Blockade dieser Funktionen. Dies weist daraufhin, dass nicht nur lösliches L1, sondern auch die zytoplasmatische Domäne von L1 funktionell aktiv ist. Im zweiten Teil der Arbeit wurde der Mechanismus, der L1-vermittelten Signaltransduktion untersucht. Die zytoplasmatische Domäne von L1 gelangt nach sequenzieller Proteolyse durch ADAM und Presenilin-abhängiger γ-Sekretase Spaltung in den Zellkern. Diese Translokation im Zusammenspiel mit der Aktivierung der MAPK ERK1/2 durch L1-Expression führt zu einer L1-abhängigen Genregulation. Die zytoplasmatische Domäne von hL1mut konnte ebenfalls im Zellkern detektiert werden, vermittelte jedoch keine Genregulation und unterdrückte die ERK1/2 Phosphorylierung. Die L1-abhängige Induktion von ERK1/2-abhängigen Genen wie Cathepsin B, β3 Integrin und IER 3 war in Zellen der L1-Mutante unterdrückt. Die Expression des Retinsäure-bindenden Proteins CRABP-II, welches in hL1 Zellen supprimiert wird, wurde in der L1-Mutante nicht verändert. Weitere biochemische Untersuchungen zeigen, dass die zytoplasmatische Domäne von L1 Komplexe mit Transkriptionsfaktoren bilden kann, die an Promoterregionen binden können. Die dargestellten Ergebnisse belegen, dass L1-Expression in Tumoren an drei Funktionen beteiligt ist; (i) L1 erhöht Zellmotilität, (ii) fördert Tumorprogression durch Hochregulation von pro-invasiven und proliferationsfördernden Genen nach Translokation in den Nukleus und (iii) schützt die Zellen mittels Regulation pro- bzw. anti-apoptotischer Gene vor Apoptose. Die mutierte Phosphorylierungsstelle im L1-Molekül ist essentiell für diese Prozesse. Die Anwendung neuer Therapien für Patienten mit L1-positiven Karzinomen kann mit Hinblick auf die guten Erfolge der Antikörper-basierenden Therapie mit dem mAk L1-11A diskutiert werden.

Molecular mechanisms controlling the precision of chemokine guided cell migration

By the end of the first day of embryonic development, zebrafish primordial germ cells (PGCs) arrive at the site where the gonad develops. In our study we investigated the mechanisms controlling the precision of primordial germ cell arrival at their target. We found that in contrast with our expectations which were based on findings in Drosophila and mouse, the endoderm does not constitute a preferred migration substrate for the PGCs. Rather, endoderm derivatives are important for later stages of organogenesis keeping the PGC clusters separated. It would be interesting to investigate the precise mechanism by which endoderm controls germ cell position in the gonad. In their migration towards the gonad, zebrafish germ cells follow the gradient of chemokine SDF-1a, which they detect using the receptor CXCR4b that is expressed on their membrane. Here we show that the C-terminal region of CXCR4b is responsible for down-regulation of receptor activity as well as for receptor internalization. We demonstrate that receptor molecules unable to internalize are less potent in guiding germ cells to the site where the gonad develops, thereby implicating chemokine receptor internalization in facilitating precision of migration during chemotaxis in vivo. We demonstrate that while CXCR4b activity positively regulates the duration of the active migration phases, the down-regulation of CXCR4b signalling by internalization limits the duration of this phase. This way, receptor signalling contributes to the persistence of germ cell migration, whereas receptor down-regulation enables the cells to stop and correct their migration path close to the target where germ cells encounter the highest chemokine signal. Chemokine receptors are involved in directing cell migration in different processes such as lymphocyte trafficking, cancer and in the development of the vascular system. The C-terminal domain of many chemokine receptors was shown to be essential for controlling receptor signalling and internalization. It would therefore be important to determine whether the role for receptor internalization in vivo as described here (allowing periodical corrections to the migration route) and the mechanisms involved (reducing the level of signalling) apply for those other events, too.

Migration and Spatial Development

This is an empirical study with theoretical interpretation and elaboration simultaneously on the migration process and the related spatial development in contemporary China. In so doing, there is always a combination of series of studies of the modernization of the migrants themselves with accumulation of forms of capital and changes of lebenswelt (life world) as well as the regions of their origins by the effective use of the gained resources from outgoing migration and remigration. With great efforts made to put the issues together for analysis, the author has taken three approaches to the study based on the political and economic institutional arrangements, the field work data and the elaboration of respective findings. First, as the analytical parts of the institutional changes, which have gone through the whole research, many of the policies from state level to townships involved in the migration, remigration and spatial development have been interpreted with Chinese political and cultural insight. The making of these, as the means of understanding the contexts of macro level and micro level cases is served as key linkages between scholarly imagination and social reality. Indeed most of the discussions made to explain the phenomena such as the sudden upsurge of migration flows, the emergence of three generations, the strong and weak trends of remigration as well as the related spatial development planning, etc are mainly due to the domination, at least the impact of governments decision-making in spite of growing market functioning in often operative manners. Secondly, case studies of the effects of migration and remigration are carried out between the years of 1995 and 2005 in the costal urban regions as designations and the interior rural regions as origins. Conducted mainly by the author, the cases drawn in the research focus on the process of migration with an accumulation of forms of capital away from home and the effective use of the resources flowing back to home areas. As a result, ways of accumulation and utilization of the economic, social and cultural capital are described and interpreted in terms of the development and modernization of both the migrants themselves and the regions where they come out from or move to in the future. Thirdly, in accordance with the findings generated from the cases, the author proposes in the final chapter an important argumentation as conclusion that the duel social-economic structure will inevitably be broken up and reformulated with flows of migrants and forms of capital they possess as types of future spatial development that will be put into practice. With scenarios and all the other conclusions worked out in the end, the research concludes that the pluralistic spatial development in the condition of constant space flows between regions can be a decisive line of thinking in the process of urbanization, industrialization and modernization in the long run in the future. Since this is an exploratory study of the past and present, the author has left some space open for academic debates and put forward suggestions on the inclusion of future research before implementing policies necessary for migration associated spatial practice and development.

Development of thermo-analytical prediction and classification models for food in thermal devices using a multi sensor system and artificial neural networks

Die thermische Verarbeitung von Lebensmitteln beeinflusst deren Qualität und ernährungsphysiologischen Eigenschaften. Im Haushalt ist die Überwachung der Temperatur innerhalb des Lebensmittels sehr schwierig. Zudem ist das Wissen über optimale Temperatur- und Zeitparameter für die verschiedenen Speisen oft unzureichend. Die optimale Steuerung der thermischen Zubereitung ist maßgeblich abhängig von der Art des Lebensmittels und der äußeren und inneren Temperatureinwirkung während des Garvorgangs. Das Ziel der Arbeiten war die Entwicklung eines automatischen Backofens, der in der Lage ist, die Art des Lebensmittels zu erkennen und die Temperatur im Inneren des Lebensmittels während des Backens zu errechnen. Die für die Temperaturberechnung benötigten Daten wurden mit mehreren Sensoren erfasst. Hierzu kam ein Infrarotthermometer, ein Infrarotabstandssensor, eine Kamera, ein Temperatursensor und ein Lambdasonde innerhalb des Ofens zum Einsatz. Ferner wurden eine Wägezelle, ein Strom- sowie Spannungs-Sensor und ein Temperatursensor außerhalb des Ofens genutzt. Die während der Aufheizphase aufgenommen Datensätze ermöglichten das Training mehrerer künstlicher neuronaler Netze, die die verschiedenen Lebensmittel in die entsprechenden Kategorien einordnen konnten, um so das optimale Backprogram auszuwählen. Zur Abschätzung der thermische Diffusivität der Nahrung, die von der Zusammensetzung (Kohlenhydrate, Fett, Protein, Wasser) abhängt, wurden mehrere künstliche neuronale Netze trainiert. Mit Ausnahme des Fettanteils der Lebensmittel konnten alle Komponenten durch verschiedene KNNs mit einem Maximum von 8 versteckten Neuronen ausreichend genau abgeschätzt werden um auf deren Grundlage die Temperatur im inneren des Lebensmittels zu berechnen. Die durchgeführte Arbeit zeigt, dass mit Hilfe verschiedenster Sensoren zur direkten beziehungsweise indirekten Messung der äußeren Eigenschaften der Lebensmittel sowie KNNs für die Kategorisierung und Abschätzung der Lebensmittelzusammensetzung die automatische Erkennung und Berechnung der inneren Temperatur von verschiedensten Lebensmitteln möglich ist.

Optimal Unsupervised Learning in Feedforward Neural Networks

We investigate the properties of feedforward neural networks trained with Hebbian learning algorithms. A new unsupervised algorithm is proposed which produces statistically uncorrelated outputs. The algorithm causes the weights of the network to converge to the eigenvectors of the input correlation with largest eigenvalues. The algorithm is closely related to the technique of Self-supervised Backpropagation, as well as other algorithms for unsupervised learning. Applications of the algorithm to texture processing, image coding, and stereo depth edge detection are given. We show that the algorithm can lead to the development of filters qualitatively similar to those found in primate visual cortex.

ADAM: A Decentralized Parallel Computer Architecture Featuring Fast Thread and Data Migration and a Uniform Hardware Abstraction

The furious pace of Moore's Law is driving computer architecture into a realm where the the speed of light is the dominant factor in system latencies. The number of clock cycles to span a chip are increasing, while the number of bits that can be accessed within a clock cycle is decreasing. Hence, it is becoming more difficult to hide latency. One alternative solution is to reduce latency by migrating threads and data, but the overhead of existing implementations has previously made migration an unserviceable solution so far. I present an architecture, implementation, and mechanisms that reduces the overhead of migration to the point where migration is a viable supplement to other latency hiding mechanisms, such as multithreading. The architecture is abstract, and presents programmers with a simple, uniform fine-grained multithreaded parallel programming model with implicit memory management. In other words, the spatial nature and implementation details (such as the number of processors) of a parallel machine are entirely hidden from the programmer. Compiler writers are encouraged to devise programming languages for the machine that guide a programmer to express their ideas in terms of objects, since objects exhibit an inherent physical locality of data and code. The machine implementation can then leverage this locality to automatically distribute data and threads across the physical machine by using a set of high performance migration mechanisms. An implementation of this architecture could migrate a null thread in 66 cycles -- over a factor of 1000 improvement over previous work. Performance also scales well; the time required to move a typical thread is only 4 to 5 times that of a null thread. Data migration performance is similar, and scales linearly with data block size. Since the performance of the migration mechanism is on par with that of an L2 cache, the implementation simulated in my work has no data caches and relies instead on multithreading and the migration mechanism to hide and reduce access latencies.

A Detailed Look at Scale and Translation Invariance in a Hierarchical Neural Model of Visual Object Recognition

The HMAX model has recently been proposed by Riesenhuber & Poggio as a hierarchical model of position- and size-invariant object recognition in visual cortex. It has also turned out to model successfully a number of other properties of the ventral visual stream (the visual pathway thought to be crucial for object recognition in cortex), and particularly of (view-tuned) neurons in macaque inferotemporal cortex, the brain area at the top of the ventral stream. The original modeling study only used ``paperclip'' stimuli, as in the corresponding physiology experiment, and did not explore systematically how model units' invariance properties depended on model parameters. In this study, we aimed at a deeper understanding of the inner workings of HMAX and its performance for various parameter settings and ``natural'' stimulus classes. We examined HMAX responses for different stimulus sizes and positions systematically and found a dependence of model units' responses on stimulus position for which a quantitative description is offered. Interestingly, we find that scale invariance properties of hierarchical neural models are not independent of stimulus class, as opposed to translation invariance, even though both are affine transformations within the image plane.

Neural Networks

We present an overview of current research on artificial neural networks, emphasizing a statistical perspective. We view neural networks as parameterized graphs that make probabilistic assumptions about data, and view learning algorithms as methods for finding parameter values that look probable in the light of the data. We discuss basic issues in representation and learning, and treat some of the practical issues that arise in fitting networks to data. We also discuss links between neural networks and the general formalism of graphical models.

Forecasting Global Temperature Variations by Neural Networks

Global temperature variations between 1861 and 1984 are forecast usingsregularization networks, multilayer perceptrons and linearsautoregression. The regularization network, optimized by stochasticsgradient descent associated with colored noise, gives the bestsforecasts. For all the models, prediction errors noticeably increasesafter 1965. These results are consistent with the hypothesis that thesclimate dynamics is characterized by low-dimensional chaos and thatsthe it may have changed at some point after 1965, which is alsosconsistent with the recent idea of climate change.s

Biologically Plausible Neural Circuits for Realization of Maximum Operations

Object recognition in the visual cortex is based on a hierarchical architecture, in which specialized brain regions along the ventral pathway extract object features of increasing levels of complexity, accompanied by greater invariance in stimulus size, position, and orientation. Recent theoretical studies postulate a non-linear pooling function, such as the maximum (MAX) operation could be fundamental in achieving such invariance. In this paper, we are concerned with neurally plausible mechanisms that may be involved in realizing the MAX operation. Four canonical circuits are proposed, each based on neural mechanisms that have been previously discussed in the context of cortical processing. Through simulations and mathematical analysis, we examine the relative performance and robustness of these mechanisms. We derive experimentally verifiable predictions for each circuit and discuss their respective physiological considerations.