In this thesis, the geometric margin is used for the first time as the robustness indicator of an uncoupled cellular neural network implementing a given Boolean function. First, robust template design for uncoupled cellular neural networks implementing linearly separable Boolean functions by support vector machines is proposed. A fast sequential minimal optimization algorithm is presented to find maximal margin classifiers, which in turn determine the robust templates. Some general properties of robust templates are investigated. An improved CFC algorithm implementing an arbitrarily given Boolean function is proposed. Two illustrative examples are provided to demonstrate the validity of the proposed method.

Under deregulated environments, transmission system is playing an increasingly vital role in the modern power markets. Despite the widespread experience of electricity restructuring during the past decade, important issues remain open about the best way to organize transmission to support reliability management and market trading. In the absence of an appropriate mechanism for the transmission services unbundling, it is unlikely to take into consideration the effects of power flows that diverge from the contract path. When individuals do not confront the respective responsibility for the transmission services, the market outcome will be inefficient. Allocations of transmission loss and costs to the market participants are essential for the transmission services unbundling. In this respect, two methods are proposed in this dissertation to provide technical and economical signals for the allocations. Based on circuit theory, the first method proposes a new procedure to trace the power flows on the transmission lines. Starting with the computation of an individual generator¡¦s or load¡¦s contributions to the line currents, the network is then decomposed into different networks. In each decomposed network, bus injections and line power flows are replaced by components related to only one generator or load. Based on the components of real power flow on the transmission lines, the allocations of transmission loss and supplementary cost can be easily performed. In the second method, AC power flow based injection shift distribution factors (ISDFs) are exploited for calculating generalized generation distribution factors (GGDFs) and generalized load distribution factors (GLDFs). GGDFs and GLDFs can be used to derive the components of the real powers on each transmission line contributed by the generators and loads. Three bus-oriented schemes are proposed to allocate the transmission loss. They are aimed to reduce the number of distribution factor computation and to reflect the activity in a competitive market. The reactive powers incurred on transmission lines due to the generators and loads are also derived using the reactive power distribution factors (RPDFs). Based on the components of complex power flow on the transmission lines, transmission cost allocation methods are proposed to charge the impacts of power transactions on the network. In order to take the service reliability into account the service charges consist of the components representing normal and the contingency states respectively.

Brain and central nervous cancer presents a significant clinical burden, accounting for 2.4% of all cancer deaths. High grade glioma is particularly deadly, with 5 year survival times of 35% or less. Traditional treatment includes tumor resection followed by radiation therapy or chemotherapy. Aggressive resection is essential in order to prolong patient life. In fact, several studies have shown that life expectancy increases with increased extent of resection. Extent of resection is burdened by the fact that surgeons must be careful not to remove functional brain tissue. Resection is incomplete more often than not due to lack of visual cues for the surgeon. He must rely on tactile sensation to distinguish tumor from healthy tissue. Methods such as intraoperative MRI and CT exist, but these require expensive equipment and special training that is not available in all surgical environments. Some laboratories have proposed small molecule dyes to solve this problem, but these are insufficient when used in an invasive tumor model. It was the goal of this research to provide an objective cue in the form of a nanoencapsulated visible dye without the need for additional equipment of changes to the surgery process itself other than injection of the dye. We hypothesized that the nanocarrier would allow staining of the tumor through passive targeting by taking advantage of the enhanced permeability and retention effect. Once the nanocarriers have reached the desired target, they would not diffuse out into healthy tissue due to their large size compared to small molecule dyes, which readily diffuse out and stain healthy tissue. To test this hypothesis, we prepared and characterized a liposomal nanocarrier encapsulating Evans blue dye. The nanocarrier was tested for safety in vitro and in vivo, then used to delineate tumor margins in an invasive rat glioma model in vivo. Microscopic analysis was then conducted to ensure only tumor tissue was stained by the nanocarrier. This thesis presents a successful method of tumor border delineation to provide surgeons with positive visual cues without the need for changes in surgical environment or techniques.

The presence of gas hydrates offshore of eastern Canada has long been inferred from estimated stability zone calculations, but the physical evidence is yet to be discovered. While geophysical evidence derived from seismic and borehole logging data provides indications of hydrate occurrence in a number of areas, the results are not regionally comprehensive and, in some cases, are inconsistent. In this study, the results of systematic seismic mapping along the Scotian and Newfoundland margins are documented. An extensive set of 2-D and 3-D, single and multi-channel, seismic reflection data comprising ~45,000 line-km was analyzed for possible evidence of hydrate. Bottom simulating reflectors (including one double BSR) were identified at five different sites, ranging between 300 and 600 m below the seafloor and in water depths of 1000 to 2900 m. The combined area of the five BSRs is 1720 km2, which comprises a small proportion of the theoretical stability zone area along the Scotian and Newfoundland margins (~635,000 km2). The apparent paucity of BSRs may relate to the rarity of gas hydrates on the margin or may be simply due to geophysical limitations in detecting hydrate.

Geochemical processes associated with gas hydrate formation lead to the growth of iron sulphides which have a geophysically-measurable magnetic signature. Detailed magnetic investigation, complemented by petrological observations, were undertaken on cores from a permafrost setting, the Mackenzie Delta (Canadian Northwest Territories) Mallik region, and two marine settings, IODP Expedition 311 cores from the Cascadia margin off Vancouver Island and the Indian National Gas Hydrate Program Expedition 1 from the Bengal Fan. Stratigraphic profiles of the fine scale variations in bulk magnetic measurements correspond to changes in lithology, grain size and pore fluid geochemistry which can be correlated on local to regional scales. The lowest values of magnetic susceptibility are observed where iron has been reduced to paramagnetic pyrite, formed in settings with high methane and sulphate or sulphide flux, such as at methane vents. High magnetic susceptibility values are observed in sediments which contain detrital magnetite, for example from glacial deposits, which has survived diagenesis. Other high magnetic susceptibility values are observed in sediments in which the ferrimagnetic iron-sulphide minerals greigite or smythite have been diagenetically introduced. These minerals are mostly found outside the sediments which host gas hydrate. The mineral textures and compositions indicate rapid disequilibrium crystallization. The unique physical and geochemical properties of the environments where gas hydrates form, including the availability of methane to fuel microbiological activity and the concentration of pore water solutes during gas hydrate formation, lead to iron sulphide precipitation from solute-rich brines. Magnetic surveying techniques help delineate anomalies related to gas hydrate deposits and the diagenesis of magnetic iron minerals related to their formation. Detailed core logging measurements and laboratory analyses of magnetic properties provide direct ties to original lithology, petrophysical properties and diagenesis caused by gas hydrate formation.

Temporal variations in gas hydrate related geochemical signatures under different deposition conditions are the primary purposes of this study. Accreted wedge located offshore Southwestern Taiwan receives high terrigenous river materials, 100 MT/yr, at present time. It is not clear how seep environment varied during the past glacial. A 25 meters long piston core was taken offshore Southwestern Taiwan on r/v Marion DuFresne. Short piston cores and box cores were also taken on r/v OR-1. Samples were analyzed for pore water dissolved sulfide, sulfate, methane, chloride, del O18, calcium, magnesium, alkalinity, pH, and sediment AVS, pyrite, inorganic carbon, del O- 18, C13. Changes in deposition environment play a major role in the study area. Three stages of geochemical processes are identified in the 25 meters long core, interchange between reduce and oxic depositional environments, with reducing condition in the top 10 m, oxic in between 10-20 meter and reducing below the 20 meter. High concentrations of dissolved sulfide, rapid sulfate depletion, increase of methane, decrease of calcium were found in pore water in the top 10 m of sediments together with high concentrations of pyrite, relatively higher proportion of coarsegrained sediment. Concentrations of pyrite were very low in sediments between 15 to 20 meters but increased rapidly from 20 to 25 meters with a maximum concentration at 400 umol/g. Chloride concentrations also increased to a maximum concentration of 630 mM at 20 m. The rapid increase of chloride indicated gas hydrate formation at this depth. Authigenic carbonate nodules were found in sediments below 20 m. The carbonate content also increased rapidly beneath this depth. Stable isotopic carbon composition of the carbonate varied rapidly beneath 20 m with a low at -28 per mil. The existence of oxic/reducing alterations indicates that methane seep may vary in the past in the study area.

The ‘giant’ Tambogrande volcanogenic massive sulphide (VMS) deposits within the Cretaceous Lancones basin of northwestern Perú are some of the largest Cu-Zn-Au-Ag-bearing massive sulphide deposits known. Limited research has been done on these deposits, hence the ore forming setting in which they developed and the key criteria that permitted such anomalous accumulation of base-metal sulphides are not understood. Based on field relationships in the host volcanic rocks and U-Pb geochronology, the deposits formed during the early stages of arc development in the latest Early Cretaceous and were related to an extensional and arc-rift phase (~105-100 Ma, phase 1). During this time, bimodal, primitive basalt-dominant volcanic rocks were erupted in a relatively deep marginal basin. Phase 1 rhyolite is tholeiitic, M-type, and considered to have formed from relatively high temperature, small batch magmas. The high heat flow and extensional setting extant during the initial stages of arc development were essential components for forming a VMS hydrothermal system. The subsequent phase 2 (~99-91 Ma) volcanic sequence comprises more evolved mafic rocks and similar, but more depleted, felsic rocks erupted in a relatively shallow marine setting. Phase 2 is interpreted to represent late-stage arc volcanism during a waning extensional regime and marked the transition to contractional tectonism. The Tambogrande deposits are particularly unusual amongst the ‘giant’ class of VMS deposits in that deposition largely occurred as seafloor mound-type and not by replacement of existing strata. Paleomorphology of the local depositional setting was defined by seafloor depressions controlled by syn-volcanic faults and rhyolitic volcanism. The depressions were the main controls on distribution and geometry of the deposits and, due to inherently confined hydrothermal venting, enhanced the efficiency of sulphide deposition. Geochemical and radiogenic isotope data indicate that the rhyolites in the VMS deposits were high temperature partial melts of the juvenile arc crust that had inherited the isotopic signatures of continental crust. Moreover, Pb isotope data suggest the metal budget was sourced almost wholly from mafic volcanic strata. Therefore, unlike the implications of many conventional models, the felsic volcanic rocks at Tambogrande are interpreted to have only played a passive role in VMS formation.

The dissertation explores the principle of proportionality as an instrument deriving from the notion of "democratic society" and applied by the European Court of Human Rights (ECtHR) in order to establish whether national institutions succeeded in striking a fair balance between the conflicting Convention rights or between competing individual and public interests. In the first chapter, after presenting the origins of the principle, the development of a three-tiered proportionality test and its dissemination, the main parameters relevant for the analysis of this principle are identified and the main issues of academic discussion concerning the application of proportionality by the ECtHR are revealed. The second chapter explores the evolution of the application of the proportionality principle in the case law of ECtHR and reveals the main features of proportionality test as applied in the early practice of Convention institutions. The third chapter deals with the interaction of the principles of proportionality and subsidiarity, in particular examining the impact of margin of appreciation doctrine upon the proportionality assessment conducted by the ECtHR and analysing the contents and interplay of the factors determining the width of the margin of appreciation. The forth chapter explores both commonalities and peculiarities of the application of proportionality principle in the context of different Convention rights and exposes the main criteria relevant to the balancing... [to full text] / Disertacijoje nagrinėjama proporcingumo principo, kaip iš demokratinės visuomenės idėjos kylančio instrumento, Europos Žmogaus Teisių Teismo (EŽTT) pasitelkiamo siekiant nustatyti, ar nacionalinės institucijos išlaikė teisingą pusiausvyrą tarp konkuruojančių Žmogaus teisių ir pagrindinių laisvių apsaugos konvencijos (Konvencijos) ginamų teisių ar tarp Konvencijos teisių ir viešų interesų, sampratos ir taikymo praktikos problematika. Pirmoje disertacijos dalyje, aptarus proporcingumo principo kilmę, trinario proporcingumo tyrimo modelio susiformavimą ir sklaidą, apsibrėžiami svarbiausi proporcingumo tyrimui reikšmingi parametrai ir atskleidžiama EŽTT taikomo proporcingumo principo sampratos problematika mokslinėje literatūroje. Antroje dalyje nagrinėjama proporcingumo principo taikymo EŽTT praktikoje raida ir plėtra, identifikuojami pagrindiniai ankstyvojoje Komisijos ir Teismo praktikoje taikyto proporcingumo testo bruožai. Trečioje dalyje tiriama proporcingumo principo sąveika su subsidiarumo principu, atskleidžiamas vertinimo laisvės doktrinos poveikis EŽTT atliekamam proporcingumo vertinimui, tiriami vertinimo laisvės apimtį lemiantys kriterijai ir jų sąveika. Ketvirtoje dalyje analizuojama proporcingumo principo taikymo praktika skirtingų Konvencijos ginamų teisių kontekste, atskleidžiant tiek bendruosius principo taikymo bruožus, tiek ir Teismo taikomos priežiūros intensyvumo nulemtus proporcingumo testo ypatumus. Paskutinėje dalyje EŽTT taikomas balansavimu grindžiamas... [toliau žr. visą tekstą]

This dissertation presents a study on velocity modeling using ocean-bottom seismometer data (OBS) collected in 2004 and 2006 on the western Scotian slope. Gas hydrate and free gas concentrations and their distribution along the Scotian margin were derived based on the velocity results modeled with two different OBS data sets. A strong velocity increase (140-300 m/s) associated with gas hydrate was modeled for a depth of 220 m below seafloor (bsf). At the base of that high velocity zone (330 mbsf) the velocity decreases with 50-130 m/s. This depth is associated with the depth of the bottom-simulating reflector (BSR) observed in previous 2-D seismic reflection data. The gas hydrate concentrations (2-18 %) based on these velocities were calculated with an effective medium model. The velocity modeling shows that a sparser OBS spacing (~ 1 km) reveals more velocity uncertainties and smaller velocity contrasts than a denser (100 m) spaced OBS array. The results of the travel-time inverse modeling are applied in a waveform inverse modeling with OBS data in the second part of the thesis. The modeling tests were performed to obtain information on OBS instrument spacings necessary to detect low-concentration gas hydrate occurrences. The model runs show that an increase in instrument spacing leads to an increasing loss of model smoothness. However, large instrument spacings (>500 m) are beneficial for covering a wide target region with only using a few instruments, but decreasing the lateral resolution limits of the subsurface targets. In general half of the instrument spacing defines the lower boundary for the lateral width of the target structure. Waveform modeling with the 2006 OBS data has shown that low frequencies (<8 Hz) in the source spectrum are necessary to recover the background velocity of the model. The starting model derived from travel-time inversion of the 2006 data is not close enough to the true model. Thus the first-arrival waveforms do not match within half a cycle. Modeling with a starting frequency of 8 Hz and and applying data with a low signal-to-noise ratio (1.25) introduces artifacts into the final model result without updating the velocity. / Graduate

In this study, a new discriminative learning framework, called soft margin estimation (SME), is proposed for estimating the parameters of continuous density hidden Markov models (HMMs). The proposed method makes direct use of the successful ideas of margin in support vector machines to improve generalization capability and decision feedback learning in discriminative training to enhance model separation in classifier design. SME directly maximizes the separation of competing models to enhance the testing samples to approach a correct decision if the deviation from training samples is within a safe margin. Frame and utterance selections are integrated into a unified framework to select the training utterances and frames critical for discriminating competing models. SME offers a flexible and rigorous framework to facilitate the incorporation of new margin-based optimization criteria into HMMs training. The choice of various loss functions is illustrated and different kinds of separation measures are defined under a unified SME framework. SME is also shown to be able to jointly optimize feature extraction and HMMs. Both the generalized probabilistic descent algorithm and the Extended Baum-Welch algorithm are applied to solve SME. SME has demonstrated its great advantage over other discriminative training methods in several speech recognition tasks. Tested on the TIDIGITS digit recognition task, the proposed SME approach achieves a string accuracy of 99.61%, the best result ever reported in literature. On the 5k-word Wall Street Journal task, SME reduced the word error rate (WER) from 5.06% of MLE models to 3.81%, with relative 25% WER reduction. This is the first attempt to show the effectiveness of margin-based acoustic modeling for large vocabulary continuous speech recognition in a HMMs framework. The generalization of SME was also well demonstrated on the Aurora 2 robust speech recognition task, with around 30% relative WER reduction from the clean-trained baseline.