This paper demonstrates how transformer language models can be improved by giving them access to relevant structured data extracted from a knowledge base. The knowledge base preparation process and modifications to transformer models are explained. We evaluate these methods on language modeling and question answering tasks. These results show that even simple additional knowledge augmentation leads to a reduction in validation loss by 73%. These methods also significantly outperform common ways of improving language models such as increasing the model size or adding more data.

The real-world sensorimotor paradigm is based on the premise that sufficient ecological complexity is a prerequisite for inducing naturally relevant sensorimotor relations in the experimental context. The aim of this thesis is to embed visual attention research within the real-world sensorimotor paradigm using an innovative mobile gaze-tracking system (EyeSeeCam, Schneider et al., 2009). Common laboratory set-ups in the field of attention research fail to create natural two-way interaction between observer and situation because they deliver pre-selected stimuli and human observer is essentially neutral or passive. EyeSeeCam, by contrast, permits an experimental design whereby the observer freely and spontaneously engages in real-world situations. By aligning a video camera in real time to the movements of the eyes, the system directly measures the observer’s perspective in a video recording and thus allows us to study vision in the context of authentic human behaviour, namely as resulting from past actions and as originating future actions. The results of this thesis demonstrate that (1) humans, when freely exploring natural environments, prefer directing their attention to local structural features of the world, (2) eyes, head and body perform distinct functions throughout this process, and (3) coordinated eye and head movements do not fully stabilize but rather continuously adjust the retinal image also during periods of quasi-stable “fixation”. These findings validate and extend the common laboratory concept of feature salience within whole-body sensorimotor actions outside the laboratory. Head and body movements roughly orient gaze, potentially driven by early stages of processing. The eyes then fine-tune the direction of gaze, potentially during higher-level stages of visual-spatial behaviour (Studies 1 and 2). Additional head-centred recordings reveal distinctive spatial biases both in the visual stimulation and the spatial allocation of gaze generated in a particular real-world situation. These spatial structures may result both from the environment and form the idiosyncrasies of the natural behaviour afforded by the situation. By contrast, when the head-centred videos are re-played as stimuli in the laboratory, gaze directions reveal a bias towards the centre of the screen. This “central bias” is likely a consequence of the laboratory set-up with its limitation to eye-in-head movements and its restricted screen (Study 3). Temporal analysis of natural visual behaviour reveals frequent synergistic interactions of eye and head that direct rather than stabilize gaze in the quasi-stable eye movement periods following saccades, leading to rich temporal dynamics of real-world retinal input (Study 4) typically not addressed in laboratory studies. Direct comparison to earlier data with respect to the visual system of cats (CatCam), frequently taken as proxy for human vision, shows that stabilizing eye movements play an even less dominant role in the natural behaviour of cats. This highlights the importance of realistic temporal dynamics of vision for models and experiments (Study 5). The approach and findings presented in this thesis demonstrate the need for and feasibility of real- world research on visual attention. Real-world paradigms permit the identification of relevant features triggered in the natural interplay between internal-physiological and external-situational sensorimotor factors. Realistic spatial and temporal characteristics of eye, head and body interactions are essential qualitative properties of reliable sensorimotor models of attention but difficult to obtain under laboratory conditions. Taken together, the data and theory presented in this thesis suggest that visual attention does not represent a pre-processing stage of object recognition but rather is an integral component of embodied action in the real world.

This study examined historic change of riparian plant communities and fluvial geomorphic response of gravel-bedded streams and their floodplains to over 50 years of hydrologic disturbance. Four tributary basins of the Arkansas River were analyzed. Lake Creek, Clear Creek, and Cottonwood Creek are drainages similar in area, physiography, and vegetation composition. However, Lake Creek may receive an instantaneous discharge of approximately 28 m3sec-1 from the Twin Lakes tunnel, over three times the normal flow of the stream during spring runoff. By contrast, Clear Creek and Cottonwood Creek, nonaugmented streams, were used as controls to compare the historic and present condition of natural flowing streams with Lake Creek. Lake Fork was also examined, another augmented stream that carries transmountain water stored in a reservoir to the Upper Arkansas River. Hydrologic data collected from U.S.G.S. gaging stations were used in an attempt to identify relations between specific flow events and corresponding changes in fluvial landforms, riparian vegetation patterns, and channel morphology. Topographic cross sections of the channel and floodplain were surveyed and hydraulic geometry at each field site was examined. Aerial photographs from 1938-39, 1956-57 , and 1988 were incorporated into a Geographic Information System (GIS) to prepare a series of land cover maps of the river corridor at eight field sites on augmented and nonaugmented streams. Results indicate channel morphology and riparian vegetation cover on the active floodplains of Lake Creek and Lake Fork have experienced substantial change from natural physical and biologic conditions as a result of historic and present flow augmentation practices. The increase in water discharge without a natural sediment load in Lake Creek and Lake Fork appears to have caused accelerated rates of channel bank erosion, incision, and bank retreat below transbasin diversion releases. In some downstream reaches characterized by gentler slopes and wider valley bottoms, the width-depth ratio of the channel has increased and sinuosity has decreased. Sites carrying transmountain water were characterized by higher shear stress and specific stream power values and larger bed material compared to control sites. Consequently, the Lake Creek and Lake Fork drainages show temporal and spatial loss of riparian vegetation adjacent to the stream channel in response to historic and present flow augmentation practices. Control watersheds, Clear Creek and Cottonwood Creek, where natural flow regimes exist, did not experience similar magnitudes of change in channel morphology or spatial distribution of riparian vegetation.

A thrust augmented nozzle for hybrid rocket systems is investigated. The design lever-ages 3-D additive manufacturing to embed a helical fuel port into the thrust chamber of a hybrid rocket burning gaseous oxygen and ABS plastic as propellants. The helical port significantly increases how quickly the fuel burns, resulting in a fuel-rich exhaust exiting the nozzle. When a secondary gaseous oxygen flow is injected into the nozzle downstream of the throat, all of the remaining unburned fuel in the plume spontaneously ignites. This secondary reaction produces additional high pressure gases that are captured by the nozzle and significantly increases the motor’s performance. Secondary injection and combustion allows a high expansion ratio (area of the nozzle exit divided by area of the throat) to be effective at low altitudes where there would normally be significantly flow separation and possibly an embedded shock wave due. The result is a 15 percent increase in produced thrust level with no loss in engine efficiency due to secondary injection. Core flow efficiency was increased significantly. Control tests performed using cylindrical fuel ports with secondary injection, and helical fuel ports without secondary injection did not exhibit this performance increase. Clearly, both the fuel-rich plume and secondary injection are essential features allowing the hybrid thrust augmentation to occur. Techniques for better design optimization are discussed.

abstract: Optimal design theory provides a general framework for the construction of experimental designs for categorical responses. For a binary response, where the possible result is one of two outcomes, the logistic regression model is widely used to relate a set of experimental factors with the probability of a positive (or negative) outcome. This research investigates and proposes alternative designs to alleviate the problem of separation in small-sample D-optimal designs for the logistic regression model. Separation causes the non-existence of maximum likelihood parameter estimates and presents a serious problem for model fitting purposes. First, it is shown that exact, multi-factor D-optimal designs for the logistic regression model can be susceptible to separation. Several logistic regression models are specified, and exact D-optimal designs of fixed sizes are constructed for each model. Sets of simulated response data are generated to estimate the probability of separation in each design. This study proves through simulation that small-sample D-optimal designs are prone to separation and that separation risk is dependent on the specified model. Additionally, it is demonstrated that exact designs of equal size constructed for the same models may have significantly different chances of encountering separation. The second portion of this research establishes an effective strategy for augmentation, where additional design runs are judiciously added to eliminate separation that has occurred in an initial design. A simulation study is used to demonstrate that augmenting runs in regions of maximum prediction variance (MPV), where the predicted probability of either response category is 50%, most reliably eliminates separation. However, it is also shown that MPV augmentation tends to yield augmented designs with lower D-efficiencies. The final portion of this research proposes a novel compound optimality criterion, DMP, that is used to construct locally optimal and robust compromise designs. A two-phase coordinate exchange algorithm is implemented to construct exact locally DMP-optimal designs. To address design dependence issues, a maximin strategy is proposed for designating a robust DMP-optimal design. A case study demonstrates that the maximin DMP-optimal design maintains comparable D-efficiencies to a corresponding Bayesian D-optimal design while offering significantly improved separation performance. / Dissertation/Thesis / Doctoral Dissertation Industrial Engineering 2019

SAMMANFATTNING Bakgrund: De flesta kvinnor ser amning som det naturligaste sättet att ge ett nyfött barn näringsrik föda. Kvinnor får information via mödravården och sjukvården att bröstmjölken innehåller den näring som ett nyfött barn behöver. Dessutom skyddar den mot några av de vanligaste barnsjukdomarna. Alla kvinnor i dagens samhälle kan inte amma. Vissa kan inte amma på grund av fysiska orsaker. Det finns också kvinnor som har genomgått bröstoperationer. I dag opererar allt fler kvinnor sina bröst av olika anledningar, exempelvis på grund av missnöje med utseendet eller ryggproblematik. Antalet operationer har ökat markant och majoriteten av kvinnorna som genomgår en bröstoperation är i reproduktiv ålder. En del kvinnor erhåller inte adekvat information angående komplikationer som kan bli bestående. Kvinnor överväger inte eventuella bieffekter som bröstkirurgi kan ha gällande deras framtida förmåga att amma. Det är viktigt för barnmorskan att så tidigt som möjligt främja amning. För att stödet från barnmorskan ska bli optimalt behövs god kunskap och en positiv attityd till amning. Syfte: Syftet med detta arbete var att beskriva vilka effekter bröstoperationer kan ha på amning och att belysa hur stöd och information kan påverka amningen efter bröstoperation. Metod: En litteraturöversikt utfördes där 23 vetenskapliga artiklar inkluderades. Artikelsökningen genomfördes i databaserna PubMed och CINAHL där artiklar publicerade mellan åren 2000 – 2014 söktes. Även manuella sökningar förekom. Artiklar både med kvalitativ och med kvantitativ ansats inkluderades i översikten. Författaren klassificerade och utförde kvalitetsbedömning på samtliga artiklar. Alla artiklar hade granskats av etisk kommitté. Resultat: Operationerna som genomförs idag med uppdaterad teknik ska inte påverka amningsförmågan om operationen utförs på rätt sätt. I studien framkom att kvinnor har olika erfarenheter av bröstoperationer. Vissa kvinnor kunde amma exklusivt, det vill säga gav sitt barn endast bröstmjölk, under en längre tid medan andra inte hade tillräckligt med mjölkmängd för att under en längre tid amma sitt barn. Det framkom också att vissa mammor inte lyckades komma igång med amningen och avstod därmed. Stödet som ges är betydande. Något som genomsyrar litteraturöversikten är den bristande informationen, det vill säga kvinnorna som opererar sig får inte adekvat information angående möjligheten att kunna amma eller eventuellt inte. Slutsats: Resultatet visade att bröstoperationer och den teknik som användes kan påverka hur vida kvinnan kan amma eller inte. Dock är inte operationen den enda avgörande faktorn för en lyckad amning. Studien som utfördes påvisade att stödet och uppmuntran som gavs till mamman hade en avgörande roll. Stödet och uppmuntran som mamman fick i samband med förlossningen och vid eventuella svårigheter var avgörande för hennes beslut om att amma eller inte.

No description available.

No description available.

This study explored whether women who have breast augmentation are more vulnerable to negative stereotypes about their intelligence and abilities than women who do not elect to have breast augmentation. Furthermore, whether the motivation for seeking such surgery contributes to stereotypes was also explored. Participants in this study read experimentally manipulated biographies of a fictitious female author, followed by her short story. The only manipulation was one line in the author's biography which varied the reason for the plastic surgery across 4 conditions (for herself, for her significant other, for medical reconstructive reasons, or no surgery). The participants were asked to read and evaluate the author and her work. The results of this study suggest that women who elect to have plastic surgery for herself or her husband are perceived as being less deep. The results also suggest that women who have previously elected to have plastic surgery judged the authors' fictional work more critically when she had plastic surgery for herself or for her husband.

Syntetiskt oxytocin är en vanlig behandling under förlossning. Oxytocin används som induktionsmetod och för att behandla värksvaghet. För att diagnostisera värksvaghet behöver kvinnan vara i aktiv fas, men varken för diagnosen värksvaghet eller för begreppet aktiv fas råder konsensus. Konsensus saknas även för längden på en normal förlossning och det råder individuell variation. Idag administreras oxytocin utan hänsyn till Body Mass Index. Syftet med litteraturöversikten var att belysa om och i så fall hur dosen av syntetiskt oxytocin påverkas av kvinnans Body Mass Index under förlossning. Metoden var en litteraturöversikt med systematisk ansats. Databassökningen utfördes i CINAHL och PubMed. Artiklarna kvalitetsgranskades och 15 kvantitativa artiklar inkluderades. Artiklarna analyserades med integrerad analys med en induktiv ansats. Dataanalysen resulterade i fyra huvudkategorier: exponering av syntetiskt oxytocin, oxytocindos, infusionslängd/duration av oxytocininfusion och varierande förlossningsutfall. Tre subkategorier påträffades under huvudkategorin oxytocindos: genomsnittlig medelhastighet, administrerad totaldos och högsta infusionshastighet. Resultatet visade att dosen av oxytocin under förlossning delvis påverkas av kvinnans Body Mass Index. Det sågs att ett högre Body Mass Index ökar sannolikheten att exponeras med oxytocin under förlossning. Ingen entydighet gällande oxytocindos i relation till Body Mass Index sågs för övriga kategorier. Det sågs en viss antydan att kvinnor som induceras behöveren högre högsta infusionshastighet och totaldos i relation till ett högre Body Mass Index. Slutsatsen var att sannolikt påverkas oxytocinadministreringen av fler faktorer än kvinnans Body Mass Index. Då oxytocin är en vanlig förlossningsintervention behövs fortsatt forskning på hur oxytocin kan doseras på ett mer individanpassat sätt. / Synthetic oxytocin is a common treatment during labor. Oxytocin is used as a method for induction and for augmentation of labor. To diagnose dystocia the woman needs to be in active stage, however there is no common definition either for the diagnose dystocia or concept of active stage. There is no agreement about the length of a normal labor and individual variety occurs. Today oxytocin is administrated without consideration for Body Mass Index. The aim of the literature review was to illustrate if and is so how the dose of synthetic oxytocin is affected by maternal Body Mass Index during labor. The method was a literature review with a systematic approach. The search was conducted in CINAHL and PubMed. The quality of the studies was assessed and 15 quantitative studies were included. The studies were analyzed through integrated analysis with an inductive approach. The data analysis resulted in four main categories: exposure of synthetic oxytocin, oxytocin dose, duration of oxytocin and varying labor outcomes. Three subcategories were found in the main category oxytocin dose: average rate, total dose and peak infusion rate. The result was that the oxytocin dose used during labor was partly affected by maternal Body Mass Index. A higher Body Mass Index was seen to increase probability for oxytocin exposure during labor. No unambiguity was seen regarding oxytocin dose in relation to Body Mass Index for the other categories. A trend towards the need of a higher peak infusion rate and total dose regarding a higher Body Mass Index was shown for women who underwent induction. In conclusion the administration of oxytocin is plausibly affected by other factors than maternal Body Mass Index. Since oxytocin is a common intervention during labor further research is needed to explore how administration of oxytocin could occur in a more individualized manner.