Treasures @ UT Dallas
Not a member yet
7697 research outputs found
Sort by
Aesthetics and Politics: Community-Based Art in Brazil
No full textThis dissertation analyzes large-scale and multifaceted artworks created in 2008 by the Dutch
artists Jeroen Koolhaas and Dre Urhahn (Haas & Hahn), the French street artist JR, and the
Brazilian mixed-media artist Vik Muniz. Each of these artists were drawn specifically to Brazil’s
second largest city, Rio de Janeiro, and decided to create large-scale works of art in informal
housing settlements known as favelas with the help of local community members. In August
2008, photographs of women’s faces and eyes from Rio de Janeiro’s oldest favela, Morro da
Providência, were enlarged and pasted on building façades and rooftops in the community,
creating a hillside of women staring out at Brazil’s Marvelous City. Two months later, Haas &
Hahn completed a nine-month-long project entitled Rio Cruzeiro, which resulted in a 2,000
square meter mural of Japanese-style carp fighting the current of a roaring river rushing down
the hillside in the favela of Vila Cruzeiro. Around this same time, Vik Muniz constructed seven
large-scale assemblages on the ground from discarded materials with the help of catadores
(pickers of recyclable materials) from the Jardim Gramacho landfill and neighboring favela,
which he photographed and entitled Pictures of Garbage.
Although the works by Haas & Hahn, JR, and Muniz I study are often understood within existing
frameworks for community-based art, I argue that this framework does not fully account for
what these works are trying to achieve in both political and aesthetic terms. Specifically, I argue
that in the community-based artworks by Haas & Hahn, JR, and Muniz in Rio de Janeiro, we see
an investment in the aesthetic and political potential of commitments not usually associated with
socially engaged art in Brazil, such as to completedness, unity, intentional meaning, and beauty.
At the same time, the works of Haas & Hahn, JR, and Muniz include in varying degrees an
element of viewer experience, ultimately defining who participates in the work and who sees the
work, as they toe the line between formalism and spectatorship. Some of the strategies seen in
the community-based artworks of Haas & Hahn, JR, and Muniz can be traced to the Concrete
and Neo-Concrete art movements in Brazil from the 1950s and 1960s. By situating these
artworks, which have been created in marginalized working-class communities in Rio de Janeiro,
in dialogue with the work of Brazilian Neoconcretists in the early 1960s, such as Lygia Clark,
Hélio Oiticica, and Lygia Pape, I reveal how they constitute a deep and continued connection
between Brazilian art and the favela. Artworks by Haas & Hahn, JR, and Vik Muniz have
incorporated aspects of the art-making process employed by Clark, Oiticica, and Pape; however,
this process yields works of art that metaphorically represent the city in which they were created.
I conclude that community-based art’s aesthetic shift from viewer experience to visual
representation in Brazil is not a departure from the social and political issues of the city, but
rather a continuation of aesthetic dialogue aimed at promoting social change in Rio de Janeiro
Design and Optimization of Behavioral Application Specific Processors
No full textMost modern Integrated Circuits (ICs) are heterogeneous System-on-Chip (SoC) that include multiple embedded processors, on-chip memories, different types of serial interfaces
and a growing number of dedicated hardware accelerators. The embedded processors are
customized to the application domain for optimal performance and energy efficiency. This
further increases the complexity of designing, optimizing and verifying the entire system.
To address this, in this thesis I have proposed different automatic design flows for the efficient
design and optimization of application specific processors (ASIPs) using High-Level Synthesis
(HLS). HLS is a promising approach that raises the level of abstraction of VLSI design from
the RT level to the behavioral level.
These flows range from building near on-chip memory processing systems to optimized embedded processors with tightly integrated hardware (HW) accelerators. In all cases, the
proposed systems lead to significant area, performance and/or power improvements by leveraging some of the main advantages of HLS like being able to efficiently share HW resources
Information Design for Online Platforms
No full textIn this dissertation we focus on strategic operations problems that arise within the technology-
enhanced platform economy. Particularly, we explore how online platforms such as Amazon,
Google and Facebook can improve their revenue performance by designing and controlling
the information flow in the marketplace. In that, taking the online platforms’ perspective, we
investigate the strategic interactions between online platforms and their market participants in
an information-decentralized environment. We consider novel, profit-maximizing, information
provision policies in this study, both in dynamic and static modeling environments. First, we
uncover an online platform’s optimal information strategy for a time-locked sales campaign,
whereby the platform selectively provides information about historical purchase decisions
to influence future customers’ product evaluation. Next, we study a practically relevant
problem on consumer privacy concerns. Particularly, in a setting where users can choose
whether and what personal preference information to share with the platform, we study how
the platform should design its profit-maximizing recommender system, that also strategically
persuades users about the relevance of recommended items. The characterization of the
platform’s optimal design of its provision policies permit us to draw important managerial
and regulatory insights
Synthesis and Bioactivity Evaluation of KDM4 Inhibitors in Prostate Cancer Cells
No full textThe need for alternative therapeutic targets has grown due to the stagnating progress of treatment
for metastatic prostate cancer with activity independent of the androgen receptor. Inhibition of
histone lysine demethylases belonging to the KDM4 subfamily are of significant interest due to
their aberrant expression and role in castration-resistant prostate cancer. This dissertation presents
the design, synthesis, and biochemical evaluation for a library of 8-hydroxyquinoline-based
derivatives of B3, a KDM4 inhibitor that has previously demonstrated therapeutic potential.
Motivated by the search for improved efficacy for KDM4 inhibition in prostate cancer, the first
investigation highlights the structure-activity relationship discovered from modifying the
phenylpropyl moiety of B3. Screening a comprehensive list of different chemical groups revealed
several with improved inhibitor activity and stability. Continuing our search for improved efficacy,
the second investigation explored augmentation of the benzamide moiety of B3 that led to the
identification of a new lead inhibitor using 4-(3-methoxypropyl)morpholine. This modification to
B3 led to the desired cytotoxicity to tumor viability during in vitro MTT assays and in vivo murine
studies
Ambulation Assessment Using Deep Learning and Fiducials
No full textOn average, hospitalized geriatric patients spend over 85% of their time in bed. Frequent
bed rest results in rapid physical deconditioning. To prevent this physical decline and the
resulting complications, it is imperative that clinical staff set and monitor ambulatory goals
with their patients. Often referred to as early and progressive ambulation, a plan to get
patients out of bed and moving as soon as possible after injury or surgery greatly reduces
the risk of additional complications and reduces hospital stays. Currently, no solution exists
to effectively monitor and report ambulation in a clinical setting, greatly impacting clinicians’
ability to monitor patient progress.
This work presents three systems for ambulation assessment based on three different technologies, real-time location systems, depth cameras, and RGB cameras. After examining the
first two technologies a in-depth look into the implementation of RGB camera-bases ambulation assessment is taken. Leveraging a custom implementation of a deep neural network-
based object detection algorithm facilitates detecting people and a set of assistive devices
relevant to clinical environments. The object detections form the basis for the quantification of different ambulatory activities and related behaviors. Combined with simple fiducial
markers and a Deep SORT algorithm, a subject can be detected and tracked simultaneously. Using features extracted from detected people and objects as input, machine learning
models are leveraged to determine a person’s ambulatory status. This status includes motion, identity, distance traveled, posture, and use of an assisted device. The summary of
an individual’s ambulatory status over time provides the data required to monitor overall
ambulatory health and progress of hospitalized patients
Essays on Blockchain Platforms: Decentralization, Delegation and Governance
No full textThis dissertation addresses economic and operational issues arising in blockchain
application by examining three decentralization mechanisms: Proof-of-work (PoW), voting
scheme, and delegation. Chapter 1 investigates the proof-of-work mechanism in the Bitcoin
system. Chapter 2 and 3 explore the vote-based mechanism and delegation features in Steemit, a
blockchain-based social media platform. The common goal in these blockchain applications is
decentralization, i.e., operating without centralized control.
The most popular blockchain application is Bitcoin. One of the key features of the Bitcoin
application is decentralization: It is designed to operate without any centralized control. This
feature is critical to the operation of the system because the system allows agents or miners to
participate in verifying transactions and record-keeping. It uses a proof-of-work (PoW) protocol
to validate transactions without the need for a trusted intermediary. The system’s decentralization
is related to the participation of miners. A large number of miners involved in the process would
enhance the security and trust of the system. In Chapter 2, we raise the question of whether the
Bitcoin system can sustain decentralization, considering the importance of decentralization. We
examine the question by drawing insights from a game-theoretical model that reflects the
participation of miners in the record-keeping process. Furthermore, we empirically support our
insights from the model using data from the Bitcoin system. We conclude that the decentralization
of the Bitcoin system is not sustainable.
Besides the PoW protocol, many blockchain platforms use other mechanisms to self-
govern. Notably, the voting system is used as one of the decision-making mechanisms for self-
governance in many blockchain applications. In Chapter 3, we primarily look into the governance
aspect of a blockchain platform that users a voting system. We examine a blockchain-based social
media platform, Steemit, which rewards its users for creating and curating blogs. The platform
uses a voting system that aggregates votes from its users and distributes rewards to its contributors
based on the votes. The platform also allows users to trade (buy or sell) votes from others to
promote their creations. We investigate whether trading votes benefit users financially and increase
their engagement in the platform. We find that vote trading positively affects users’ rewards and
engagement in the short run. In the long run, users’ productivity reduces, and the platform sees
fewer contributions. Moreover, we provide short-run and long-run implications of trading votes
using Steemit’s blockchain transactions data.
Chapter 4 mainly focuses on the delegation feature in Steemit. The current trend among
blockchain applications shows that delegation has become a prominent feature, specifically for
decentralized autonomous organizations (DAOs). It is related to transferring decision-making
power like voting rights to others. In the traditional investment world, delegating investment
decisions to financial experts is quite common. Individual investors rationally delegate the
responsibility to financial managers for their expertise and seek better returns from their
investment. However, delegating investments (stakes) to other users in peer-to-peer platforms may
not have similar implications as delegating investment decisions to professionals, specifically
when the platforms are decentralized. We investigate how the delegation feature helps users and
the platforms. We find that delegation brings higher rewards to users. However, delegating stakes
or voting rights reduces users’ engagement. Our findings have many implications for platform
designers, researchers, and policymakers. We believe that the lack of thoughtful design of the
decentralized mechanisms could negatively impact the quality of decision-making in a
decentralized organization. Ultimately, these mechanisms can affect the long-term sustainability
of the platform
Investigation on the Organization of Turbulence for High Reynolds-number Boundary-layers Through LiDAR Experiments
No full textLight Detection And Ranging (LiDAR) technology has gained growing attention for research
in the realm of atmospheric turbulence due to its capability to probe the atmospheric boundary layer (ABL) with high spatial and temporal resolution within a volume with height and
horizontal extent comparable to the ABL thickness. In this work, several high Reynolds-
number turbulent flows, such as ABL for onshore and marine environment, wakes generated
by utility-scale wind turbines, have been probed with the LiDAR anemometry with the
aim of investigating the variability of the mean kinetic energy, the spatial and spectral heterogeneity of the streamwise momentum resulting from the dynamics of various turbulent
eddies. In the first part of this dissertation, the LiDAR spatial averaging process, which is the
source for a reduced turbulence intensity measured though a wind LiDAR, has been systematically corrected through a novel data-driven procedure based on the quantification of the
energy damping for the streamwise velocity spectra at high-frequencies owing to the inertial
sub-range. This correction method has enabled reverting the low-pass filtering operated by
the LiDAR measuring system on the turbulent velocity fluctuations, and obtain a corrected
estimate of the second-order statistical moment of the streamwise velocity. Subsequently,
LiDAR measurements collected for the ABL evolving over a very flat and homogeneous terrain are interrogated to investigate the distribution of the streamwise turbulence intensity
associated with wall-attached eddies and larger coherent structures as a function of height.
This work has culminated with the proposition of an analytical model for the prediction of
the linear coherence spectrum, which is based on the Townsend’s attached eddy hypothesis.
Finally, LiDAR experiments performed for marine ABL, ABL interacting with utility-scale
wind turbines, and coupling LiDAR with snow particle image velocimetry to investigate
atmospheric turbulence are presented
Rumor Source Detection: an Application in Social Network Analysis
No full textMisinformation spreads quickly on social media. Identifying the source of rumors is crucial
for controlling their spread, similar to finding patient zero in an epidemic. The aim of rumor
Blocking (RB) problem is to design algorithms and mechanisms for controlling rumor diffusion
in social networks. Identifying the rumor sources is one of the most popular research topics
of rumor blocking in the field of social network analytics. The first part of the research
dives deep into the existing studies on social network analysis and its applications extended
further to specifically focus on the RB problem. Then we study the problem of identifying a
single rumor source based on the observation of the state of the network (whether nodes are
infected by the rumor or not) by proposing a greedy algorithm to find the best k monitoring
stations among all rumor-infected nodes with a 2-approximation. This method is quite good,
but it has a limitation: it only considers the structure of the network, not the content of
the rumor or the behaviour of the users spreading it. To address this, we introduce a new
data-driven approach to identify multiple rumor sources in social networks leveraging Graph
Neural Networks (GNNs), a cutting-edge technology for analyzing complex networks
A Large-eddy Simulation Study of Surface Layer Response to Roughness Spatial Heterogeneity
No full textA rough wall is known to modify the structures of the turbulent boundary layer relative to
the boundary layer flow over a smooth wall due to eddies, vortices, and turbulence structures
introduced by drag-exerting surface irregularities. The elevated turbulence and mixing due
to roughness influence momentum and heat exchange. The roughness also affects flow separation
and reattachment and may result in secondary flow. The present study examines surface
layer response to roughness heterogeneity in the streamwise and spanwise directions, resulting
from various arrangements of drag-exerting tree canopy elements. Previous studies have
shown that a predominant streamwise-aligned spanwise heterogeneity results in undulations
in streamwise velocity in the transverse-wall-normal plane where low-momentum pathways
(LMP) and high-momentum pathways (HMP) are located based on surface roughness characteristics,
known as Prandtl’s secondary flow of the second kind. Several numerical and
experimental studies reported an HMP over the elevated roughness, whereas others witnessed
an LMP over high roughness flanked by counter-rotating vortices corresponding with momentum
excess. In addition, flow response to a step change in surface roughness in the flow-wise
direction is also widely studied. Through a suite of controlled numerical experiments using
large-eddy simulations (LES), we examined the confluence of roughness heterogeneity on the
generation and sustenance of secondary flow to address the disagreement above. The roughness is introduced as synthetic trees using the canopy drag model. The results showed that
within the transitional roughness regime in the spanwise direction, 1 ≲ δ2/δ ≲ 2, where δ2 is
the gap between streamwise aligned parallel rows of canopies and δ is the channel half height,
an upwelling of low-momentum fluid is observed over canopy rows for cases with δ1/h ≲ 6
(d-type roughness), where δ1 is the downstream gap between subsequent trees and h is the
canopy height from the ground. However, a downwelling of high momentum fluid aloft is
found for δ1/h ≳ 6 (k-type roughness) (Raupach and Rajagopalan, 1991; Chung et al., 2021).
The observations suggest that the roughness regimes (d and k-type) regulate the polarity of
the secondary cells. The flow response learned from simulations categorized the roughness
regimes on a δ1 − δ2 parameter space. This suggests that the upwelling and downwelling of
low and high-momentum fluid over the elevated roughness depends on the streamwise cavity
width. Within the topography regime, δ2/δ ≳ 2, both d- and k-type roughness cases result
in the downwelling of high momentum fluid above canopy elements. Later, the transitional
roughness regime resulting from δ-scale spanwise roughness heterogeneity is explored to understand
the flow response near critical cavity width, δ1/h ≈ 6. We show that the system
attains a large-scale non-periodic reversal of the secondary flow near the critical streamwise
gap using statistical measures such as joint probability distribution, time series evolution,
and probability density functions
Essays on Labor Issues in Accounting
No full textThis dissertation consists of two parts. In the first part, I document the economywide extent of
corporate violations related to non-financial stakeholders such as employees, environment, and
customers and the associated incentives provided in executive bonus plans. Using firm-level data
on corporate violations from 47 federal agencies, I find that the number of violations per year is
lower in firms with bonus plans that place less emphasis on expenses, a practice referred to as
“cost-shielding”, compared to those with cost-cutting bonus plans. The relation between costshielding incentives and corporate violations is more pronounced for firms with weaker corporate
governance, firms that operate in competitive industries, and firms with less monopsony power. In
additional tests, I confirm my findings using more detailed data from the U.S. Department of
Labor’s Wage and Hour Division and document that the number of employment-related violations
decreases after managers receive an employee-related goal in their bonus plans for the first time.
These findings highlight a novel link between incentives that stem from compensation plans and
corporate compliance practices –an indicator of the risks borne by firms to non-financial
stakeholders. In the second part, I examine the effect of the change in employee incentives on the
extent of accounting restatements using the adoption of Employee Stock Ownership Plans
(ESOPs) as an exogenous shock. ESOPs provide employees with long-term ownership in the firm
by investing the firm’s stock in employees’ retirement accounts. I find evidence of an increased
probability of restating financial statements after the adoption of ESOPs. Overall, this study
contributes to our understanding of the forces that affect accounting restatements