Universiẗat, Diss., 2001--Bonn.

Decades have passed since the realisation that a protein’s amino acid sequence can contain all the information required to form a complex three-dimensional fold. Until recently, these encoded structures were thought to be crucial determinants of protein function. Much effort was directed to fully understand the mechanisms behind how and why proteins fold, with natively unfolded proteins thought to be experimental artefacts. Today, the field of natively unfolded – or so-called intrinsically disordered – proteins, is rapidly developing. Protein disorder content has been positively correlated with organismal complexity, with over thirty percent of eukaryotic proteins predicted to contain disordered regions. However, the biophysical consequences of disorder are yet to be fully determined. With the aim of addressing some of the outstanding questions, the work described in this thesis focuses on the relevance of structure within disordered proteins. Whilst populating a variety of conformations in isolation, a subset of disordered proteins can fold upon binding to a partner macromolecule. This folded state may be present within the ensemble of conformations sampled by the unbound protein, opening the question of what comes first: folding or binding? Protein engineering techniques were employed to alter the level of residual ‘bound-like’ structure within the free conformational ensemble, and the consequences on coupled folding and binding reactions were investigated. Resultant changes in the rate of association are easily imaginable; yet, this work demonstrates that the majority of the observed changes in binding affinity were due to alterations in the rate of dissociation, thus altering the lifetime of the bound complex. Promiscuous binding is a touted advantage of being disordered. If many disordered proteins, each with their own conformational ensemble, can bind and fold to the same partner, then where is the folding component encoded? Does the partner protein template the folding reaction? Or, is the folding information contained within the disordered protein sequence? Utilising phi-value analysis on the BCL-2 family of proteins, residues in the disordered sequence were probed to ascertain which form contacts at the transition state of the reaction. Comparison with phi-value analyses of alternative pairs – sharing either the ordered or disordered protein – provides insight into the encoding of these interactions. In the context of a bimolecular reaction, the amino acid sequence of the disordered protein was shown to determine the interactions within the transition state. Thus, analogous to the discovery from decades’ past, it is the sequence of the protein that folds which encodes its pathway, even when binding is a prerequisite.

This thesis will address the problem of the parallel evaluation of algorithmic trading models based on multiple kernel support vector regression. Various approaches to parallelization of the evaluation of these models will be proposed and their suitability for highly parallel architectures, namely the Intel Xeon Phi coprocessor, will be analysed considering specifics of this coprocessor and also specifics of its programming. Based on this analysis a prototype will be implemented, and its performance will be compared to a serial and multi-core baseline pursuant to executed experiments. Powered by TCPDF (www.tcpdf.org)

Two analyses of data recorded in proton-proton collisions at the ATLAS detector in 2012 are presented in this thesis. The first pertains to the beam separation (van der Meer) scans required to calibrate the absolute luminosity. An estimate is made for the size of the correction needed to the standard van der Meer calibration method, which assumes that the proton bunch density profiles are factorisable. This is done by observing and modelling the evolution of various beam spot phenomena during the separation scans. The second analysis described is a series of measurements of the Z/gamma cross-section, differential in the phi star observable, for different ranges of the boson invariant mass and absolute rapidity. In particular the events in which the boson decays to electron-positron pairs are considered. The phi star observable is defined in terms of the well-measured lepton directions and enables a probe of initial state gluon radiation in the non-perturbative regime of QCD.

This thesis presents phenomenological studies of a state-of-the-art NNLL+NLO theoretical calculation of a novel collider observable known as 'phi star'. In these studies the 'phi star' observable, a measure of azimuthal decorrelation, is applied directly to the leptons in the production of massive lepton pairs in hadron collisions (the Drell-Yan process). This provides an alternate measure of the recoil of the massive vector boson (Z/gamma) against initial state QCD radiation, but with distinct experimental advantages over the traditional boson transverse momentum. Attention is focused on the small-'phi star' regime (the quasi-back-to-back regime) where the infrared dynamics of soft/collinear gluon emissions become important. These phenomenological studies are followed up with the presentation of a measurement of 'phi star' in 'Z to mu mu' events using 20.3 fb^-1 of collision data collected by the ATLAS experiment in 2012. Finally, studies directly related to the ATLAS absolute luminosity calibration by the van der Meer (vdM) method are presented, with the objective of elucidating the role of transverse linear beam correlation. In particular, I present studies using an analytical method I have developed in order to precisely extract individual beam information by way of studying phenomena pertaining to the luminous region during vdM scans. In addition, a dedicated study of the long- and short-term stabilities of the principal detectors for luminosity monitoring is also presented, along with an appropriate recalibration of these detectors.

This dissertation describes detailed performance engineering and optimization of an unstructured computational aerodynamics software system with irregular memory accesses on various multi- and many-core emerging high performance computing scalable architectures, which are expected to be the building blocks of energy-austere exascale systems, and on which algorithmic- and architecture-oriented optimizations are essential for achieving worthy performance. We investigate several state-of-the-practice shared-memory optimization techniques applied to key kernels for the important problem class of unstructured meshes. We illustrate for a broad spectrum of emerging microprocessor architectures as representatives of the compute units in contemporary leading supercomputers, identifying and addressing performance challenges without compromising the floating-point numerics of the original code. While the linear algebraic kernels are bottlenecked by memory bandwidth for even modest numbers of hardware cores sharing a common address space, the edge-based loop kernels, which arise in the control volume discretization of the conservation law residuals and in the formation of the preconditioner for the Jacobian by finite-differencing the conservation law residuals, are compute-intensive and effectively exploit contemporary multi- and many-core processing hardware. We therefore employ low- and high-level algorithmic- and architecture-specific code optimizations and tuning in light of thread- and data-level parallelism, with a focus on strong thread scaling at the node-level. Our approaches are based upon novel multi-level hierarchical workload distribution mechanisms of data across different compute units (from the address space down to the registers) within every hardware core. We analyze the demonstrated aerodynamics application on specific computing architectures to develop certain performance metrics and models to bespeak the upper and lower bounds of the performance. We present significant full application speedup relative to the baseline code, on a succession of many-core processor architectures, i.e., Intel Xeon Phi Knights Corner (5.0x) and Knights Landing (2.9x). In addition, the performance of Knights Landing outperforms, at significantly lower power consumption, Intel Xeon Skylake with nearly twofold speedup. These optimizations are expected to be of value for many other unstructured mesh partial differential equation-based scientific applications as multi- and many- core architecture evolves.

abstract: Many core modern multiprocessor systems-on-chip offers tremendous power and performance optimization opportunities by tuning thousands of potential voltage, frequency and core configurations. Applications running on these architectures are becoming increasingly complex. As the basic building blocks, which make up the application, change during runtime, different configurations may become optimal with respect to power, performance or other metrics. Identifying the optimal configuration at runtime is a daunting task due to a large number of workloads and configurations. Therefore, there is a strong need to evaluate the metrics of interest as a function of the supported configurations. This thesis focuses on two different types of modern multiprocessor systems-on-chip (SoC): Mobile heterogeneous systems and tile based Intel Xeon Phi architecture. For mobile heterogeneous systems, this thesis presents a novel methodology that can accurately instrument different types of applications with specific performance monitoring calls. These calls provide a rich set of performance statistics at a basic block level while the application runs on the target platform. The target architecture used for this work (Odroid XU3) is capable of running at 4940 different frequency and core combinations. With the help of instrumented application vast amount of characterization data is collected that provides details about performance, power and CPU state at every instrumented basic block across 19 different types of applications. The vast amount of data collected has enabled two runtime schemes. The first work provides a methodology to find optimal configurations in heterogeneous architecture using classifiers and demonstrates an average increase of 93%, 81% and 6% in performance per watt compared to the interactive, ondemand and powersave governors, respectively. The second work using same data shows a novel imitation learning framework for dynamically controlling the type, number, and the frequencies of active cores to achieve an average of 109% PPW improvement compared to the default governors. This work also presents how to accurately profile tile based Intel Xeon Phi architecture while training different types of neural networks using open image dataset on deep learning framework. The data collected allows deep exploratory analysis. It also showcases how different hardware parameters affect performance of Xeon Phi. / Dissertation/Thesis / Masters Thesis Engineering 2019

The odds ratio is one of the most widely used measures of association for 2 × 2 tables. A generalized Yule coefficient transforms the odds ratio into a correlation-like scale with a range from -1 to 1. Yule's Y, Yule's Q, Digby's H, and a new coefficient are special cases of a generalized Yule coefficient. The new coefficient is shown to be similar in value to the phi coefficient. A confidence interval and sample size formula for a generalized Yule coefficient are proposed. The proposed confidence interval is shown to perform much better than the Wald intervals that are implemented in statistical packages.

Incremental Phi* och Field D* är vägplaneringsalgoritmer som uppfyller två egenskaper. Den första är att deras vägar kan korsa en miljö i vilken vinkel som helst. Den andra är att de i en okänd miljö kan planera om sina vägar snabbt ifall de stöter på ett hinder. Detta kan vara användbart i realtidsstrategispel. Detta arbete testar därför deras förmåga att skapa korta vägar och att planera om vägar snabbt för att ta reda på deras styrkor och svagheter. Utöver detta testas också deras tider i den första sökningen de gör när miljön är outforskad samt i de fall där algoritmerna har visat sig ha svagheter, vilket är i tomma miljöer och i återvändsgränder. Resultatet är att Incremental Phi* hittar kortare vägar och planerar om vägar snabbare. Den får också kortare tider i en tom miljö, medan Field D* får bättre tider i den första sökningen och i återvändsgränder. / <p>Det finns övrigt digitalt material (t.ex. film-, bild- eller ljudfiler) eller modeller/artefakter tillhörande examensarbetet som ska skickas till arkivet.</p><p>There are other digital material (eg film, image or audio files) or models/artifacts that belongs to the thesis and need to be archived.</p>

Community colleges are in a unique position in the hierarchy of higher education. They provide quality education at an affordable price. They are strategically located in communities to provide educational opportunities for everyone. The latest prediction is that enrollment at two-year institutions is expected to increase from 5.7 million students to 6.3 million students by 2012 (Gerald & Hussar, 2002). However, research indicates that the retention and transfer rates of community college students are low. This is problematic due to the increased calls for accountability on the national education agenda which are linked to student outcomes. This is astounding for a system with a foundational belief in self-development and an unquenchable mission to provide postsecondary access to people who would likely not attend college if such avenues did not exist. Fortunately, there is one aspect of this statistic that brings renewed optimism to the university transfer mission of community colleges. Eightyive percent of Phi Theta Kappa members transfer to a four-year institution. Currently more than 600 colleges and universities in all 50 states, the District of Columbia, Canada, and London offer more than $36 million in transfer scholarships to Phi Theta Kappa members (www.ptk.org). This study examined the experiences of those students who are members of Phi Theta Kappa in an effort to identify ways to recruit and retain students and increase the transfer percentage for community college students. The Community College Student Experiences Questionnaire was used to examine their community college experiences quantitatively. Descriptive statistics and Spearman correlations were used to analyze the data. Results of this study indicated that the Phi Theta Kappans were attending their community college to prepare for transfer to a four-year college or university. As anticipated, the Phi Theta Kappans were more involved in activities related to their courses and computer technology. The Quality of Effort scale indicated a low level of engagement in cultural activities. An unexpected finding was that there was no statistical significant difference in student satisfaction based on age, sex, or gender. Additionally, the students were satisfied with their community college environment.