Alliance One Tobacco (Malawi)
Academic Research Repository at the Institute of Developing EconomiesNot a member yet
52305 research outputs found
Sort by
Towards A Programmable Nanomechanical Interface for Mediating Spin-Spin Interactions
No full textSolid state spin qubits are promising candidates for quantum information processing. In particular, the nitrogen vacancy (NV) center in diamond is known to have coherence times exceeding milliseconds even at room temperature. However, due to the limits of qubit fabrication and the short-range nature of magnetic dipolar interactions, it remains difficult to generate programmable interactions between a large number of NV centers. To address this challenge, it has been proposed to use nanomechanical resonators as a mesoscopic interface between solid state spin qubits. In this thesis, I will describe experimental efforts in building a scanning probe platform, where individual NV centers in diamond nanopillars are coupled to magnetially functionalized silicon nitride mechanical resonators. The scanning probe configuration enables programmable connectivity via mechanical transport of the nanopillars. Proof-of-principle measurements show that the coherence of the NV center is preserved despite relative movement in a magnetic field gradient, by utilizing the nitrogen nuclear spin as a quantum memory. I will also describe measurements of the spin-mechanical coupling via both DC and AC magnetometry. Finally, I present some preliminary results related to sensing of a single NV center with the mechanical resonator, which demonstrate the high level of control over each subsystem. With realistic improvements to several system parameters, high spin-mechanical cooperativities are feasible, offering a new avenue towards scalable quantum information processing with spin qubits.Physic
Safe Optimistic Concurrency in Modern C++: A Reimplementation of Masstree
No full textThe memory model introduced in C++11 marked a significant advancement in writing consistent and correct concurrent programs across diverse hardware platforms. However, concerns remain that this progress in portability and safety may have come at the cost of performance. In this thesis, we investigate the performance implications of using C++11-standard concurrency primitives in the context of high-performance data structures.
We begin with an overview of the C++11 memory model, focusing on the header and its utilities. We then review recent developments in concurrent key-value stores, with particular attention to their use of synchronization mechanisms.
Our primary contribution is a reimplementation of Masstree, a high-performance, tree-based key-value store. This reimplementation is designed to avoid all forms of undefined behavior under the C++11 memory model. This is achieved in spite of Masstree’s reliance on complex, optimistic concurrency techniques. Our implementation closely replicates the behavior of the original beta release by the Masstree authors, which depended on non-standard behaviors for performance.
Through analysis of disassembled binaries and performance benchmarks on x86-64 systems, we demonstrate that adopting standard-compliant C++11 concurrency features incurs minimal performance overhead. Our findings suggest that careful engineering using standard tools can yield both correctness and efficiency in concurrent system design.Computer Scienc
To Heat or Eat: Current Patterns of Energy Poverty and Redlining in the United States
No full textRedlining, a federal housing policy that began in 1934 and continued until 1968, continues to shape economic and environmental injustices across the United States. This housing policy restricted minority groups and people of color from accruing the most common form of social and monetary equity and wealth over much of the mid-20th century. It is possible that people in these groups currently face a higher energy burden, defined as the percentage of gross household income spent on energy costs (National Renewable Energy Laboratory, 2022). The objective of my research was to understand the relationship between the current spatial distribution of energy burden and the historical redlining process in the residential housing sector across different cities in the United States. The two major questions I addressed were: What is the pattern between current instances of energy burden and places where redlining was applied? And based on the statistical regions identified by the U.S. Census Bureau (e.g. Northeast, Midwest, etc.), what are the regional differences in these patterns when comparing energy burden vs temperature?
Related to question 1, I hypothesized that the chance of being energy burdened was higher in formerly redlined communities compared to the three other respective classes that comprise the non-redlined communities. I further hypothesized that the northern regions of the United States have higher variances in the instances of energy burden when comparing formerly redlined communities to non-redlined communities within a region, due to the longer and colder winters in these areas compared to the southern regions. A new method for dataset development was used, followed by statistical analysis to identify the potential patterns between redlined areas and rates of energy burden.
The results showed that census tracts located within formerly redlined communities have higher instances of energy burden, with the lowest energy burdens in communities that were graded A, and slight increases in energy burden for the remaining grades C, B and D. This trend was consistent for all U.S. statistical regions. Cities in the South showed the widest range of energy burden values for communities formerly graded at C and D. The research results could be used in policymaking to justify programs targeting formerly redlined areas to alleviate energy burden through weatherization and energy affordability programs from the state and federal level.Extension Studie
Image-Based Algorithms for Remote Surgical Site Infection Diagnosis in Rural Rwanda
No full textA key challenge in global health is the delivery of high-quality health care in low-resource settings. In sub-Saharan Africa, wound infections are physically and financially catastrophic for patients due to limited access to medical facilities, health professionals, and adequate follow-up care. Surgical Site Infections (SSIs) in rural Rwanda affect an average of 12.4% of women undergoing cesarean deliveries and significantly contribute to maternal morbidity and mortality.
A key challenge to SSI monitoring is that cesarean-associated SSIs often develop after hospital discharge and traveling back to a facility for follow-up care is burdensome for women. However, the widespread availability of mobile health technology among community health workers (CHWs) in rural regions presents a new opportunity for SSI care. In this thesis, we propose a CHW-led mobile health system for automated, home-based diagnosis of surgical site infections for mothers after C-section. The mobile health system takes a smartphone image of the wound site, uses a real-time computer vision algorithm for image preprocessing, and deploys an AI-based predictive model for image classification. The methodology eliminates the need for tedious manual preprocessing and the variability in the collected image data.
The utility of our method is demonstrated by testing on two image datasets -- visible and thermal -- of 621 wound photos collected from women at Kirehe District Hospital in Rwanda. The final model trained on visible images had an AUC=0.86 (sensitivity=0.83, specificity=0.75). The final model trained on the thermal images had an AUC=0.89 (sensitivity=0.94, specificity=0.81).
Overall, we demonstrate the first smartphone-based methodology for surgical site infection diagnosis. Further, the thermal model provides a new opportunity to help identify wound infections without dependence on skin color. We believe this new diagnostic tool will reduce barriers to accessing post-partum follow-up care, and takes a promising first step toward improving maternal health in low-resource settings. Further, we hope this advancement paves the way for the practical implementation of image-based machine learning models and represents an important milestone for the adoption of artificial intelligence for global health.Computer Scienc
Foreign Firms in Vietnam and the US–China Trade War
No full textapplication/pdfIDP000951_001This study empirically investigates ownership of foreign firms in Vietnam during the US–China trade war. In our empirical analysis, we identify firm’s nationality by two indicators: the country of the largest shareholder or investor, and the directors’ nationalities. This differentiation plays a key role in identifying so-called “Singapore-washing” in Chinese firms. Our findings can be summarized as follows. First, foreign direct investment from Singapore is mostly conducted by non-Singaporeans. However, there are only a small number of Chinese directors’ firms with an investment source from Singapore. Second, among firms in Vietnam with Chinese directors, those that entered after the start of the US–China trade war or those with an investment source from countries other than China have a lower propensity for trade. Third, the larger presence of firms with an investment from China is associated with higher export growth to the US, but firms with Chinese directors are not.technical repor
