Treasures @ UT Dallas
Not a member yet
7697 research outputs found
Sort by
Study in Big Data Harnessing and Related Problems
No full textSocial networks, such as Facebook and Twitter, have provided incredible opportunities for
social communication between web users around the world. Social network analysis is an
important problem in data harnessing. The analysis of social networks helps summarizing
the interests and opinions of users (nodes), discovering patterns from the interactions (edges)
between users, and mining the events that take place in online platforms. The information
obtained by analyzing social networks could be especially valuable for many applications.
Some typical examples include online advertisement targeting, viral marketing, personalized
recommendation, health social media, social influence analysis, and citation network analysis.
In this dissertation, we study two types of applications emerging from modern online social
platforms in the view of social influence. One is influence maximization(IM) problem from
a discount-based online viral marketing scenario, which aims at maximizing influence in the
adoption of target products, and the other is online rumor source detection problem, in which
the spread of misinformation is supposed to be minimized and the source is expected to be
detected. We formulate them as set function optimization problems and design solutions with
performance guarantees. In study of set function optimization, there is a challenge coming
from the submodularity of objective function. That is, some of the practical problems are
not submodular or supermodular, the existing greedy strategy cannot be directly applied
to problems to get a guaranteed approximate solution. To solve those non-submodular and
non-supermodular problems, one method called DS decomposition has been considered, in
which given a set function, we decompose it to be representable as a difference between
submodular functions. Based on this method, we further study a problem about how to find
a DS decomposition efficiently and effectively. Then we propose a generalized framework
that is made up of our novel algorithms under deterministic version and random version
respectively to solve maximization of DS decomposition and show their performances under
various combinatorial settings. In addition, we discuss our findings on the role of black-box,
that has been an important component in study of computational complexity theory as well
as has been used for establishing the hardness of problems, about its implied power and
limitations in study of data-driven computation for proving solutions to some computational
problems
Multi-scale Modeling of Dislocation-driven Plasticity in Sub-micron Scale Metals
No full textMetals are the cornerstone for industrial and structural applications due to their strength and
ductility. To meet the constantly increasing requirements of advancing technology, a superior
combination of strength and ductility is necessary. With this motivation, materials research has
discovered novel methods to manipulate the internal microstructure leading to changes in the
mechanical behavior. Since the plastic deformation in metals is commonly mediated by the motion
and interaction of crystalline defects, it is vital to attain an in-depth understanding of how these
defect microstructures could affect the macroscopic properties and how they can be manipulated
for optimal performance.
Interestingly, as the sample size approaches microstructural dimensions, crystalline metals start to
show size dependence in their mechanical properties, deviating from the bulk behavior where
mechanical response is size independent. At the micron and sub-micron length scales, plastic
deformation is mainly driven by individual dislocations rather than the stochastic interaction
between dislocations, so that it is imperative to delve into the detailed mechanics of individual
dislocations for a fundamental understanding of microstructure-property relations. Dislocation
dynamics (DD) simulations can be a useful tool due to their capability to keep track of the detailed
dislocation microstructure evolution and to predict the corresponding macroscopic material
response. In this regard, the DD framework along with theoretical approaches are utilized to study
the correlation between the size dependent mechanical response and the dislocation microstructure
at the micron and sub-micron length scales.
In the first part of this dissertation, the plasticity in body-centered cubic (BCC) metals is examined
with particular focus on the strong temperature dependence, which stems from the thermally
activated motion of screw dislocations. In Chapter 2, we develop a DD model based on the
atomistic characterization of dislocation mobility and potential source mechanisms to investigate
the temperature dependent plasticity in BCC micropillars. Our models of molybdenum (Mo) and
niobium (Nb) show the dislocation source mechanism changes with respect to temperature due to
the change in mobility of screw dislocations, and these results are compared with experimental
results and theoretical models. In addition, the size dependence increases with temperature, which
agrees with recent experimental observations.
To understand the underlying mechanisms that control the mechanical properties of nanostructured
metals, an insight into the role of the grain boundary in dislocation-driven plastic deformation is
vital. The grain boundary has been observed as a dislocation source, sink, or having no effect,
which in turn, gives rise to different macroscopic mechanical responses. With this motivation, the
second part of this dissertation (Chapter 3) describes the atomistic simulations and threedimensional DD simulations that were performed to investigate dislocation interactions at various
grain boundaries and their role in the plastic deformation of face-centered cubic (FCC)
bicrystalline micropillars. The atomistically-informed DD simulations show that bicrystalline
samples containing a high angle grain boundary (HAGB) display hardening and higher flow
stresses compared to single crystals, while micropillars with a coherent twin boundary (CTB) show
similar flow stresses to the reference single crystalline samples. This is due to the transparency of
the grain boundary to slip transmission, which is observed in the atomistic simulations.
Interestingly, allowing dislocation glide on the grain boundary exhibits a decrease in flow stress
as slip transmission becomes easier.
To further investigate deformation mechanisms, it is necessary to delve into the detailed motion
of individual defects under complex loading conditions. Current dislocation dynamics modeling
techniques are limited to fixed geometries and small deformation. To overcome these limitations,
a newly developed multi-scale model called the defect dynamics element method (DDEM) could
be used, as it couples dislocation dynamics with a finite element model. In the third part of this
dissertation, Chapter 4, this new coupled model is applied to model the Taylor impact test of BCC
tantalum (Ta) single crystals, and investigates the anisotropic mechanical behavior observed in
recent experimental findings. Finally, a summary of the completed work is provided in Chapter 5
Pharmacological Regulation of Protein Translation in Fragile X Syndrome
No full textThe behavioral hallmarks of Autism Spectrum Disorder (ASD) are driven by molecular
mechanisms that remain largely unknown. One model of ASD involves the deletion of the
Fmr1 gene. Mutations in Fmr1 cause Fragile X Syndrome (FXS) and are the most
common monogenetic source of intellectual disability in humans. The Fmr1 gene encodes
for Fragile X Mental Retardation Protein (FMRP). FMRP is a conserved RNA-binding
protein that binds the ribosome and attenuates translation. Loss of FMRP results in
aberrant protein translation. This genetic lesion results in widespread cognitive
deficiencies, specifically with learning, memory, and social interaction. A deeper
understanding of the mechanism driving FMRP-dependent synaptic plasticity enables
identification of highly specific therapeutics that ameliorate core neurological deficits
associated with the disorder. In FXS, eIF4E is hyperphosphorylated and has emerged as
a therapeutic target. Yet, specific inhibitors of Mitogen-Activated Protein Kinase
Interacting Protein Kinase (MNK) have not been examined in the context of Fmr1 -/y mice.
The goal of this research is to understand if compensation of FMRP loss can be achieved
through manipulation of translation by the MNK-eIF4E regulatory axis. The results of this
dissertation outline the viability of a novel therapeutic for the reversal of behaviors
associated with Fragile X Syndrome, for which there are no FDA-approved treatments
Large Eddy Simulation Study of Turbulent Flow During Transport of Aeolian Material
No full textIn order to understand aeolian transportation, it is important to understand the origin of
the aeolian source materials. The wind-driven hopping motion of sand called saltation is
the dominating form of aeolian transport. When saltating particles splash on the sediment
bed, they release fine dust. Aerodynamic surface stress imposed by the mixing processes of
the inertia-dominated surface layer drives saltation. Due to the high spatial and temporal
variability of the imposed stress, saltation itself becomes intermittent. Saltation is initiated
when the surface shear velocity exceeds the fluid threshold of the sand particles. During
saltation, splashing sand grains release fine dust known as aerosols. Saltation sustains itself
with the continuation of the splashing of sand grains. This process stops when the shear
velocity falls below the impact threshold value. So sand mobilization starts with the exceedance of the fluid threshold and continues until the shear value falls below the impact
threshold. Even when the time-averaged shear velocity is below the fluid threshold, saltation
can still be started due to turbulent fluctuations. This also accounts for transport intermittency which is difficult to capture via transport models. We utilized the dynamic properties
of large-eddy-simulations to record a time series of the shear velocity imposed by the flow.
This time series was later conditionally sampled to filter out the frequencies where saltation
was active. These time series were used to recover probability density functions from which
a compensated shear value was calculated. This compensated subthreshold shear velocity
exhibited a linear relationship with the actual time-averaged shear velocity. A new transport
model based on the compensated subthreshold shear velocity was proposed that performed
well against field data measurements. Atmospheric surface layer flows are known to exhibit
alternative patches of long sinuous structures of momentum excess and deficit. While the
former is associated with negative vertical velocities, the latter is associated with positive
vertical velocities. The high stress beneath these high momentum structures (HMRs) initiate
saltation but the downwash from aloft traps the dust in the saltation layer itself and inhibits
dust suspension. The low momentum structures (LMRs) are capable of dust suspension but
they barely start sand mobilization. This paradox poses a challenge to explain the entrainment mechanisms. The existence of the paradox was established by analyzing the data from
field measurements. These HMRS and LMRs were investigated with the help of quadrant
analysis and classified into four categories: sweeps, ejections, inner and outer interactions.
These four turbulent production mechanisms were investigated along with the hysteresis
phenomena to propose three possible dust entrainment pathways. Numerically these dust
entrainment modes were accounted for by conditional sampling techniques. The numerical
results substantiated the proposed modes of dust entrainment. The density normalized conditions of the LES simulation make the results applicable to both terrestrial and Martian
conditions. The credibility of the simulation results are established with a grid insensitivity
test conducted with three different mesh densities
The Role of Relevant Peptides on the Sex-specific Nature of Migraine
No full textMigraine is considered to be both one of the most disabling and common disorders worldwide yet,
migraine is both more common and disabling among women. While this disorder effects 2-3 times
as many women as men across the globe the reasons for this disparity are not known. These data
present new insights regarding how calcitonin gene-related peptide interacts with the
predominantly female hormone prolactin to create female-specific hypersensitivity in rodent
models of migraine. This sensitivity to dural CGRP is also contingent on the presence of ovarian
secreted hormones, although it is not yet clear which are responsible for mediating this effect.
While CGRP is likely in part responsible for the prevalence of migraine in women, in this
dissertation we present evidence that dural amylin leads to hypersensitivity in males, but not in
females. As such amylin may mediate migraine in males; however, ovariectomized females
demonstrate more robust responses to dural amylin, than their male counterparts. This suggests
that ovarian derived hormones may block responses to dural amylin. The data covered in this
project may provide valuable information for the development of sex-specific migraine
therapeutics
RNA Control in Sensory Neurons: a Functional Genomics Approach
No full textThis work investigates the molecular mechanisms of pain signaling at the level of translation in
sensory neurons. Pain-sensing neurons, or nociceptors, are integral to the genesis of many forms
of chronic pain. Here, we apply functional genomics approaches to examine the role of
translational control in pain. High-throughput sequencing experiments reveal potential molecular
targets, which are confirmed using pharmacology, electrophysiology, and behavioral methods.
With these methods, we examine the role of the nonsense mediated decay (NMD) pathway,
eukaryotic initiation factor 5A (eIF5A), and the S6 ribosomal protein kinase 1 (S6K1) in pain.
Along with these findings, we investigate the use of human induced pluripotent stem cell
(hiPSC)-derived sensory neurons in pain research using high-throughput RNA sequencing. We
also present one of the few datasets employing ribosome profiling on the dorsal root ganglion
(DRG) in the presence of inflammatory mediators. Lastly, we show our foundational work
linking NMD to pain. Overall, these findings highlight the role of translational control in pain
Comparing the Prediction Value of Different Measures of Subjective Memory Complaints on Structural Brain Differences in Healthy Aging
No full textPrevious work has investigated the validity of Subjective Memory (SM) complaints in predicting
structural changes associated with aging with mixed results. The current dissertation aimed to
further the understanding of the association between structural changes and SM and investigate
the utility of SM discrepancy variables both in cross sectional and a longitudinal approach. Two
SM discrepancy variables, which examined the discrepancy between perceived and actual
memory failures, were calculated two ways to determine which measure would be more sensitive
in predicting white matter hyperintensities and cortical thickness in a cross-sectional sample of
(N = 110 participants). We found no relationship between SM or either SM discrepancy variable
when predicting cross sectional CTh but did find that the standard measure of SM and the two
discrepancy scores predicted higher WMH burden in the frontal and parietal lobes. Next, we
examined how change in CTh across a four-year lag related to our three SM variables at baseline
and at follow up using a smaller sample (N = 60) and found that SM score and SM discrepancy
at baseline predicted CTh change in frontal, temporal, and parietal regions. In contrast, only the
standard MFQ was predictive of CTh decline in the frontal, parietal, and temporal regions. These
findings support previous research that SM complaints relate to more than just behavioral
differences and may represent differences relating to early sings of structural changes associated
with aging. Additionally, the sensitivity differences when comparing the standard SM score to
the discrepancy scores depends on the time of assessment and structural component being
examined. Further research including more impaired individuals should be investigated to
determine if a pattern with cross sectional CTh differences exists. Further, how would
impairment change the relationship with the three SM variables in a longitudinal setting
The Politics of a Game Patch: Patch Note Documents and the Patching Processes in League of Legends
No full textThis thesis traces the histories of the digital game League of Legends through the game’s
patching processes and the documentation about these patches written by the game’s developers,
Riot Games. Specifically, by analyzing the game’s histories of racial representation and
professional player labor through a focus on the effects of Riot Games’ patching processes, this
thesis investigates the politics of a game patch in constituting a post-racial logic and precarious
labor practices for professional players and beyond. In this investigation, I begin by examining
the introduction of racial and sexual diversity game patches during the process of League of
Legends’s rise to global popularity. By considering the introduction of characters that are
explicitly coded as racially and sexually diverse, such as Caitlyn, Lucian, and Neeko, alongside
the inclusion of microtransaction cosmetic appearances that alter the racial coding of player
characters, I argue that Riot Games reproduces the logics of racial othering in these patches
through a positioning of these differences as merely aesthetic preferences, which collectively
contribute towards League of Legends’ status as a globally palatable game that aestheticizes race
and reinscribes a post-racial logic into the game’s universe. Building on this, I proceed to
analyze the patching processes and consequences surrounding the professional gaming event
Worlds 2015, a prestigious international tournament. Building on accounts of professional
players alongside the journalistic entries on Worlds 2015, I connect the tenuous and precarious
labor conditions of professional eSports players at the whims of developer patches with scholarly
theorizations on the processes of technological obsolescence and decay. In so doing, I argue that
patches are far from its conception as a strictly reactive process. Instead, patches and patching
are locations whereby developer goals are actively negotiated with players. To conclude, this
thesis investigation ends with a story of the rise and fall of a localized League of Legends’
competitive scene that connects the post-racial logic and precarious labor practices in game
patches. By problematizing how race and labor are intimately tethered to the processes of game
patching, I highlight how patching processes can reproduce ludo-Orientalist logic in reality
Algorithms to Compute Discrete Residues of a Rational Function
No full textThe classical notion of residue, for a rational function with complex coefficients, is a powerful
and ubiquitous tool, having applications in many different areas. For example: Complex
Analysis, Physics, Number Theory, Differential Equations, and Combinatorics, to name
a few. In the last decade several new notions of discrete residues have been developed by
different researchers, all of which have in common the following obstruction-theoretic feature:
a given rational function f (x) is “special” (e.g., rationally integrable, or rationally summable,
or rationally q-summable) if and only if all of its corresponding residues are zero. All of these
notions of residue (both the classical one and also its discrete variants) are originally defined
in terms of a complete partial fraction decomposition of the given rational function f (x),
which is too expensive to carry out in practice due to the high computational cost of finding
the complete factorization of the denominator. The main contribution of this dissertation is
the development of an efficient factorization-free algorithm to compute the discrete residues
of a rational function
Copy Ahead Segment Ring: an Ephemeral Memtable Design for Distributed LSM Tree
No full textWe present a novel Memtable design for distributed LSM trees that effectively reduces contention between concurrent range scans, writes, and prune operations. This design emphasizes the centralization of the Memtable on Disaggregated Memory to eliminate internode
replication and reduce associated costs. Our approach can be adapted to include fixed-size
page allocation, which helps minimize memory fragmentation within Disaggregated Memory. Our approach builds upon established methodologies like Copy on Write BTree, Hashed
Wheel Timer, Separation of Key & Value, and Single Thread per Partition with Lock Free
Ring Buffer to effectively reduce contention and improve system throughput. By integrating
these proven methodologies into our approach, we have developed a highly effective solution
that leverages extra memory to achieve better throughput under contention. Overall, our
approach represents a significant step towards exploring the potential of large disaggregated
memory to improve the performance of distributed LSM tree databases