Filter Results
62568 results
Over the course of his career, the artist Rembrandt van Rijn produced eight prints depicting St. Jerome, a fourth-century ascetic and theologian. In this thesis, I examine why Rembrandt chose Jerome to represent so often in print and in what ways the saint’s imagery and iconography was adapted to reflect seventeenth-century Dutch culture and values. Rembrandt’s prints of Jerome can be divided into two even groups, with four etchings of the saint as repentant sinner and four of the saint as scholar. In the prints of the repentant Jerome, I argue that Rembrandt focuses on the act of prayer and quiet contemplation to align with Protestant ideas of penitence. In this way, Rembrandt reflects a Protestant version of asceticism that could be used to model prayer and penitence in the daily life of Dutch citizens worried for the fate of their souls. One of these four prints of Jerome in penitence was produced in collaboration with a printmaker, Jan van Vliet. With this print, I analyze the practical reasons surrounding Rembrandt’s production and posit that his collaboration with Van Vliet broadly and the prints of Jerome specifically were meant to advertise his skill, attract patrons, and attain success on the market. I lastly discuss Rembrandt’s four prints of Jerome as scholar and view them through the lens of Christian humanism and intellectual culture prevalent in the Dutch Republic during this time. Jerome’s status as patron saint of scholars and advocate for the study of languages and primacy of the original Word resonated with Protestant thinkers of the early modern era. The characterization of Jerome’s scholarship again aligned with a newly developing form of asceticism that focused on virtue and engagement with God. This thesis argues that all eight prints of Jerome in his dual guises as sinner and scholar reveal Rembrandt’s construction of the saint as an ideal of religious devotion. In these prints, Jerome’s penitence and scholarship both provide a model for the path to virtue as believers sought a way to be reassured of their salvation and lead a life of pious humility.
Data Types:
  • Other
For the optimization of production in an operating marginal oil field, it is necessary to consider the reservoir inflow, the artificial lift systems, as well as the surface facilities. Since most reservoir simulation software does not include detailed facility modeling, an integrated model of an entire field has been developed including the surface facilities, to allow detailed modeling of the entire field operation. This model is useful for optimizing production and for use in field surveillance activities, as well as investigating the applicability of simplified engineering assumptions
Data Types:
  • Other
Statistical inference, such as hypothesis testing and calculating a confidence interval, is an important tool for accessing uncertainty in machine learning and statistical problems. Stochastic gradient methods, such as stochastic gradient descent (SGD), have recently been successfully applied to point estimation in large scale machine learning problems. In this work, we present novel stochastic gradient methods for statistical inference in large scale machine learning problems. Unregularized M -estimation using SGD. Using SGD with a fixed step size, we demonstrate that the average of such SGD sequences can be used for statistical inference, after proper scaling. An intuitive analysis using the Ornstein-Uhlenbeck process suggests that such averages are asymptotically normal. From a practical perspective, our SGD-based inference procedure is a first order method, and is well-suited for large scale problems. To show its merits, we apply it to both synthetic and real datasets, and demonstrate that its accuracy is comparable to classical statistical methods, while requiring potentially far less computation. Approximate Newton-based statistical inference using only stochastic gradients for unregularized M -estimation. We present a novel inference framework for convex empirical risk minimization, using approximate stochastic Newton steps. The proposed algorithm is based on the notion of finite differences and allows the approximation of a Hessian-vector product from first-order information. In theory, our method efficiently computes the statistical error covariance in M -estimation for unregularized convex learning problems, without using exact second order information, or resampling the entire data set. In practice, we demonstrate the effectiveness of our framework on large-scale machine learning problems, that go even beyond convexity: as a highlight, our work can be used to detect certain adversarial attacks on neural networks. High dimensional linear regression statistical inference using only stochastic gra- dients. As an extension of the approximate Newton-based statistical inference algorithm for unregularized problems, we present a similar algorithm, using only stochastic gradients, for statistical inference in high dimensional linear regression, where the number of features is much larger than the number of samples. Stochastic gradient methods for time series analysis. We present a novel stochastic gradient descent algorithm for time series analysis, which correctly captures correlation structures in a time series dataset during optimization. Instead of uniformly sampling indices in vanilla SGD, we uniformly sample contiguous blocks of indices, where the block length depends on the dataset.
Data Types:
  • Other
River deltas are densely populated and dynamically changing environments located at the boundary between land and sea. Population demands for water as well as rising sea levels are increasingly threatening aquifer water quality in deltaic regions. The rate at which aquifer contamination by salt water or other contaminants occurs is dictated, in part, by the arrangement of sediment within the subsurface. In this work, we examine the heterogeneity of the subsurface from a structural vantage to better understand how surface processes and geometry are linked to subsurface architecture. The numerical model, DeltaRCM, is applied to simulate delta evolution under a variety of input conditions. The resulting model outputs simulate 800 years during which the growing delta generates a subsurface volume that is over 40m deep. Surface channel properties and behavior, such as channel depths and channel planform decay rates are measured. Similarly, the structure of the sand bodies in the subsurface domain is evaluated. These different types of analyses, surface and subsurface, are ultimately compared to take a first-look at how channel properties in a deltaic environment may relate to subsurface structure and form. Broadly, expectations about channel trends and subsurface structure from the field of geomorphology are supported. Channel depths decrease with distance from the inlet, and as the input sand proportion increases. Similarly, the channelized fraction of the delta surface increases with higher input sand fraction values. In the subsurface, different types of channel behavior on the surface correspond to different structures. The sand bodies are larger when the surface channels are shallower and more mobile. In addition, the spatial continuity within strike sections (sections taken perpendicular to the inlet channel) increases with channel depth. Comparisons of the modeled subsurface with stochastically re-arranged replicates have confirmed the assertion that surface processes create unique subsurface structures. When the input proportion of sediment contains at least 40% sand by volume, the average size of the subsurface sand bodies follows a power-law relation with respect to surface channel depths and the average channelized fraction of the delta platform. The range of spatial entropy (disorder) also increases with channel depth. Within models, with increasing distance from the inlet both channel depths and spatial entropy ranges decrease. Changing the input sediment proportions over the course of the delta evolution provides mixed results. Some channel parameters like channel depth are indistinguishable from steady input cases, while others are influenced by the initial topographic setup. In the subsurface, variable sediment input proportions create vastly different sand body geometries depending on the rate of variation of the input sand proportion. When the input sand proportion is gradually increased, the average sand body size becomes very large; however when the sand input is abruptly increased, the mean sand body value is less than a steady sand input analog.
Data Types:
  • Other
Aria for contrabass and piano / Tetsuo Kawakami -- Sonata for double bass and piano / Paul Hindemith -- A Carmen fantasy for double bass and piano / Frank Proto
Data Types:
  • Other
The traffic assignment problem (TAP) represents the final and most computationally difficult step in the Urban Transportation Modeling System (UTMS). Prior work has studied several means of solving this problem quickly. This thesis furthers our understanding of this problem by investigating two methods designed to speed up TAP solution scenarios. A free and open-source implementation of the UTMS was implemented from scratch, including an implementation of Dial’s Algorithm B to solve the TAP. Chapter II of this thesis investigates the effects on computation time of a single Algorithm B parameter: the number of equilibrations performed on solution bushes prior to improvement. The results show that determining the best-performing value for this parameter is network-dependent, and that a point exists beyond which additional equilibrations do not provide improvements in runtime, potentially slowing down some networks’ computation. Chapter III investigates methods of approximating TAP solutions using artificial neural networks (ANNs) in the context of the network design problem, in which multiple different network designs may be tested. To remove the need to re-solve the TAP under each scenario, an ANN is trained to predict link flows given changes in network capacity. The results show that this method provides close approximations to the analytical solution in substantially less time than evaluating all possible scenarios from scratch.
Data Types:
  • Other
Sonatine modale : pour flute et clarinette, op. 155 / Charles Koechlin -- Histoire du tango / Astor Piazzolla -- Sonata for flute, viola and harp, L. 137 / Claude Debussy -- Assobio a jato = The jet whistle / Heitor Villa-Lobos.
Data Types:
  • Other
This dissertation focuses on the use of precious metals in ultra-high vacuum to get a better mechanistic understanding of industrially-relevant processes. Ethanol (EtOH) is an excellent candidate as an alternative energy source because the infrastructure is already constructed to effectively distribute it as a liquid and because processes are continuously improving to make EtOH from non-edible biomass. Therefore, we explored the dehydrogenation, decomposition, and oxidation of EtOH on a Pd-Au alloy surface in ultra-high vacuum using quadrupole mass spectrometry, Auger electron spectroscopy, and reflection-absorption infrared spectroscopy coupled with support from theoretical calculations. With growing interest in a hydrogen fuel economy, dehydrogenation from EtOH can serve as an alternative to produce H₂. With molecular beam experiments and DFT calculations, we correlated mechanistic differences of EtOH dehydrogenation with changes in the Pd ensemble size on the Au(111) substrate. Next, we investigated the decomposition of EtOH and observed preferential C-C bond breakage over that of the C-O on our Pd-Au catalysts. A major hindrance to direct ethanol fuel cells is incomplete dehydrogenation, oftentimes yielding acetaldehyde as an unwanted byproduct. Therefore, gaining insight on how EtOH decomposes is beneficial to designing better catalysts. Additionally, we investigated the oxidative self-esterification of ethanol to produce ethyl acetate. Here, we observed how the presence of oxygen and hydroxyl group enables facile EtOH dehydrogenation to acetaldehyde, enabling cross-coupling to form an ester. In whole, the compilation of my work is an example of how mechanistic understanding in controlled environments can lead to better catalytic design for real-world applications.
Data Types:
  • Other
What is the relationship between public opinion and public policy? This question is at the heart of representative democracy. This dissertation attempts to enhance our understanding of the role that partisanship plays in the opinion-policy process. We proceed in four steps. First, section 1 situates the analyses that follow in the current literature. Section 2 uses data on spending preferences to estimate general spending preferences of individuals and congressional candidates in a shared dimension. The approach employed allows for direct comparison between those two groups, and between the groups and where they perceive policy to be. Section 3 investigates whether partisans respond to policy changes similarly. Findings indicate that partisans react differently to policy change in issue areas with relatively large disagreement. Finally, Section 4 flips the equation and considers policy as the dependent variable. Are partisans more likely to get their preferred policies when they control the White House? The answer, it seems, is yes. Policy responds primarily to partisans of the same party as the president.
Data Types:
  • Other
This year-long study explores the critical literacies of the Youth Equity Agents, who were youth of color in a youth participatory action research (YPAR) project and class at Community High School (CHS). The dual purposes of this study were to partner with youth investigating personally important social justice issues and to produce new knowledge about engaging critical literacies through YPAR in school. This study draws on theories of critical literacy (Comber, 2001; Freire & Macedo, 1987) and multiliteracies (Cope & Kalantzis, 2009; New London Group, 1996) to analyze the critical and other literacy practices of youth, as well as the pedagogies that engaged those practices in a school class. Through critical ethnography (Carspecken, 1996) of the project, this study shows that the Youth Equity Agents had both love for and concerns about CHS. To understand their experiences, they engaged in “reading school,” a practice similar to Freire and Macedo’s (1987) concept of critical literacy as reading the world. A critical pedagogical approach facilitated an inquiry community where youth concerns about racial equity and education could be addressed in school, as did the epistemological approach of YPAR, which values youth knowledge production (Cammarota & Fine, 2008; Irizarry & Brown, 2014; Mirra et al., 2015; Morrell, 2004). Through intra-active pedagogies (Lenz-Taguchi, 2011) youth’s critical literacies were extended in encounters with the material environments of schools. Drawing on this material learning, youth “wrote school” by authoring counterstories (Delgado, 1989) of CHS and other urban schools, as well their visions for justice in the design of a new CHS. Across their work in the project youth engaged a wide array of literacies including talk, movement, and photography. This study has implications for understanding critical literacies and YPAR, as well as instruction in those areas. It illuminates how material, embodied, and social aspects of critical literacies are woven into everyday life, but can also be engaged pedagogically. Finally, this study argues for the potential of engaging critical literacies through in-school YPAR, but also raises questions about the labor experienced by youth of color engaging in this critical work.
Data Types:
  • Other