71446 research outputs found
Sort by
Analysis of Capacity-Building Gaps and Strategic Recommendations for Strengthening National Malaria Programs in High-Burden to High-Impact Countries.
No full textBackground: Malaria remains a significant public health challenge worldwide, with approximately 70% of the disease burden concentrated in 11 High-Burden to High-Impact (HBHI) countries, as identified by the World Health Organization (WHO). Notwithstanding considerable investments and advancements in malaria interventions, such as Insecticide-Treated Nets (ITNs), rapid diagnostic tests (RDTs) and Artemisinin-based Combination Therapy (ACT), Indoor Residual Spraying (IRS), Seasonal Malaria Chemoprevention (SMC), and the recent introduction of vaccines, persistent Human Resource (HR) capacity deficiencies considerably undermine the effectiveness and sustainability of malaria control initiatives. Pre-service training and continuing education are often sub-optimal for Community Health Workers (CHWs), district health officials, and national program managers for diagnostic procedures, case management, prevention, community engagement, data analytics, and program implementation, particularly in remote and high-transmission environments.
Methods: This analysis employed a mixed-methods approach that integrated a review of peer-reviewed literature, reports from the WHO and donor agencies, and case studies from HBHI countries. Data from national health information systems, such as District Health Information System 2 (DHIS2), and malaria program reports were compiled and evaluated using an Impact-Feasibility Matrix. Systemic relationships were mapped across the community, district, and national levels. At the same time, recommendations were substantiated through consultations with technical experts, Malaria Policy Advisory Group (MPAG) members, managers of National Malaria Control Programs (NMCP), and partner organizations.
Results: Significant training and workforce shortages were identified at all tiers of the health system. At the community level, deficiencies in CHW training and supervision were correlated with instances of misdiagnosis, treatment delays, and low levels of community engagement. District-level challenges included suboptimal supply chain management, insufficient surveillance mechanisms, and limited managerial capabilities. At the national level, constraints involved inadequate expertise in research, surveillance, data analysis for policy adaptation and prioritization for subnational tailoring, and leadership. To mitigate these issues, a blended learning strategy that encompasses self-paced digital modules, virtual mentoring, and Outreach Training and Supportive Supervision (OTSS) is recommended. An application of Kirkpatrick’s four-level evaluation framework of outcomes was also recommended to assess the impacts of the proposed training strategies that would enhance learner satisfaction, knowledge acquisition, practical application, and the overall health outcomes. The Training Resource Hub, created by Malaria Eradication Scientific Alliance (MESA), has been strengthened by the inclusion of data from this study to provide an inventory of blended malaria-related training opportunities to malaria program leaders worldwide.
Conclusions: Enhancing the workforce capacity for malaria control through blended learning methodologies can effectively bridge critical HR gaps in HBHI countries. By incorporating context-specific training, digital resources, and sustainable mentorship frameworks, malaria programs can significantly improve operational efficiency, service delivery, and strategic planning. Many of these strategies have already demonstrated benefits at the local or regional level, and the wider application of lessons learned is essential if malaria elimination targets in resilient health systems are to be achieved. With the threat of an imminent decrease in donor funding, it is more important than ever that local capacity is strengthened for implementation of best practice at all levels of the health care system.Author's Origina
Challenges in assessing the impacts of regulation of Artificial Intelligence
Get PDFThe recent surge in Generative Artificial Intelligence has introduced both opportunities and risks to society. This paper discusses the challenges in assessing the impacts of regulation of AI. It identifies a range of different concerns that might give rise to AI regulation and sets out approaches that may inform the design of AI regulation as well as principles for a robust AI regulatory framework.
The paper focuses on the methodologies and challenges involved in evaluating the impacts of AI regulation particularly where there is both significant uncertainty around the costs and benefits of the proposed regulation and the potential for near-existential risk, meaning that AI regulatory proposals are not easily susceptible to standard cost benefit analysis approaches. It outlines and considers the use of a range of quantitative and qualitative approaches to the assessment of AI regulatory proposals including breakeven analysis, using real options and applying the precautionary principle.
Given the potential for significant and near-existential harm from AI, it seems reasonable and appropriate that policymakers should err on the side of caution in designing AI regulation in line with the precautionary principle. However, there are important insights from both the approach adopted for assessing environmental regulations in terms of developing a standardized metric of regulatory risk, developing more robust qualitative reasoning, and also considering the regulatory framework as a real option, implying that policymakers should retain flexibility and monitor developments in designing regulations as new evidence becomes available. This effectively views AI regulation as an investment with embedded real options (to delay, expand, revise or abandon). It requires ongoing monitoring of the effectiveness (or otherwise) of regulation and implications of wider developments in the AI space, as well as a willingness to re-open regulatory decisions in the light of new information.Version of Recor
FL* theory of the pseudogap, and the transition to d-wave superconductivity in the cuprates
Get PDFPhysicsAuthor's Origina
Arrangements Containing Shapes: Mathematical Features and their Use in Visual Calculating
No full textConstruction lines and registration marks in shape grammars ground the appearance of shapes to provide an algorithmic approach to visual calculating. Here, the two concepts are studied in unity as a new object called point-line arrangement. This chapter develops such topics as the comparison and classification of shapes by their arrangements, the characterization of the “geometries” that arrangements give rise to, the algebra of arrangements, and others. In this way, it provides a more holistic view of construction lines and registration marks, beyond the usual roles they receive in algorithmic implementations of shape grammars.Computer ScienceHarvard Data Science InitiativeAccepted Manuscrip
Combatting Collusion Between Reinforcement Learning Agents in Electricity Markets
No full textWhen markets are well behaved, we expect firms to produce at the point where marginal revenue matches marginal cost. Collusive behavior, on the other hand, arises when firms produce less than this, leading to elevated prices, lower social welfare and higher industry profits.
It is interesting, then, that collusive behavior has been observed between reinforcement learning (RL) agents that act to set prices for goods across repeated interactions in simple, simulated markets. This behavior is the convergence toward a market equilibrium that has lower social welfare or higher industry profits than what is considered a Nash equilibrium for the reinforcement learning agents. In this project, I create a simplified model of an electricity market to confirm the collusive behavior of RL agents, comparing theoretical baselines of profit and welfare to the result of using Q-Learning agents. I then study the effect of various market interventions, in both this simplified model and Abada and Lambin’s model \cite{Abada-Lambin}. The interventions I consider include a) the introduction of a welfare-maximizing agent, b) setting limits on battery and output capacity, c) the use of taxation, and d) a reward-punishment scheme.
In order to assess the suitability of each intervention, a game-theoretic equilibrium is calculated for each intervention and compared to theoretical baselines. This is computed using quadratic program solvers and Scipy optimization packages. The intervention is then implemented in an OpenAI Gym environment to confirm or reject the game-theoretic improvements that were demonstrated. For the welfare-maximizing agent intervention, it was also implemented on the Abada and Lambin model to explore how agents react to the intervention in a more complex environment.
A first result, in both the simplified model as well as Abada and Lambin’s model, is that the introduction of a welfare-maximizing agent fails to provide a desired improvement in social welfare. Likewise, creating restrictions on battery and output capacity fails to provide a desired improvement in social welfare. Rather, I show that a promising direction is to make use of a suitable taxation or reward-punishment scheme, with this able to improve social welfare in both models.Applied Mathematic
Robot Dividends: Private Sector AI Investment Dynamics in the US
No full textThroughout human history, there have been several breakthrough general-purpose technologies that have momentous implications for society and greatly expand the realm of what human beings can accomplish. These general-purpose technologies include inventions like farming, the factory system, the development of materials like iron and bronze, printing presses, electricity and the internet. However, in aggregate, there have not been that many general-purpose technology breakthroughs with one major study pegging the number at twenty-four (Suleyman et al., 2023). Strong evidence suggests that Artificial Intelligence (“AI”) is a general-purpose technology. Although the term Artificial Intelligence was first coined in 1956 and there have been several periods with reduced funding in AI research, notably the AI winters of 1974 – 1980 (Muthukrishnan et. Al, 2020) and 1987 – 1994 (Werner, 2024), over the past few years AI has become a critical area of development for corporate and state actors.
This rapid development of AI, especially in the US, is due in part to the current technological backdrop. Key characteristics of this backdrop include broad digitization, increased computing power and the ability to create new solutions by combining existing technologies (Brynjolfsson et al, 2014). As a result, AI can now recognize faces, drive vehicles, compose music, interact with customers, write computer code and ultimately solve complex problems while teaching itself new skills. Much of the funding for AI development is coming from Big Tech corporations given their vast resources and considering the unprecedented pace of AI advancements, these corporations are under intense competitive pressures to invest in AI in order to protect their market positioning. This competitive tension between corporations extends to the state level with an increasing focus on geopolitical considerations being associated with AI development, which effectively translates to subsidies for Big Tech company investment into AI. An example of Big Tech subsidies resulting from perceived geopolitical tensions is the CHIPS and Science Act where US semiconductor companies received ~$53 billion of government funding starting in 2022 (White House Fact Sheet, 2024). Competitive pressures with regards to AI development have contributed to the deprioritization of human capital and AI investment has often come at the expense of employment within the US. In addition to escalating competitive pressures, several factors including current tax and fiscal policy associated with a concentration of power within the US (derived from structural considerations) along with a lack of an overall federal framework contribute to inertia within the US government to address the possible negative impact on the labor market resulting from private sector AI investment. This thesis analyzes recent data to explore the link between private sector AI investment and employment stability in the US and argues that the private sector in the US is not incentivized to invest in AI to help promote stable employment and consequently proposes potential structural reforms to address this dynamic.Extension Studie
The Impact of Welfare on Inter-group Relations
No full textWhat explains the persistence of inter-ethnic divisions, and how can they be reduced by government policy? I argue that such divisions persist, in part, because people rely on their co-ethnics as a safety net in times of need.
Around the world, religious and ethnic groups help their members cope with shocks to their lives and livelihoods, ranging from mutual aid societies of immigrant ethnic groups in nineteenth-century USA to sectarian networks of credit and social services in Lebanon. I theorize that ethnic groups are especially well-suited to serving as a basis for social insurance because they leverage shared norms of reciprocity and solidarity to overcome problems that plague any non-state insurance arrangement, such as information asymmetry, adverse selection, and moral hazard.
But support from co-ethnics is not costless. Prior to asking their co-ethnics for help in times of need, individuals invest scarce time and resources on maintaining ties with co-ethnics, prioritize supporting co-ethnics over non-coethnics, and adhere to group-based social norms. This greater social investment in co-ethnics limits the possibility of making similar investments in ties with non-coethnics. Individuals could share many social and economic interests with non-coethnics, and ties with non-coethnics could be beneficial. Reliance on the ethnic group for insurance inhibits the formation of such inter-group ties. Ethnicity-based social insurance, therefore, reinforces inter-group divisions.
The theory implies that when the welfare state provides individuals with an alternative source of economic support, it reduces the extent to which they depend on their ethnic group as a safety net. In doing so, welfare allows individuals to form productive ties with non-coethnics, increasing inter-ethnic integration. I test this implication in the context of caste-based networks in India, focusing on an income support program for farmers in the state of Telangana. Launched in 2018, the Rythu Bandhu Scheme (RBS) provides residents of Telangana owning agricultural land with Rs. 10,000 (USD 125) per acre per year. The transfers are timed with sowing season when farmers typically have the greatest need for funds to meet consumption and investments needs.
Panel data on household loans reveal that RBS reduced the likelihood of borrowing from caste members by 38.5%. Using original survey data from 3,020 households in a difference-in-differences framework, I show that inter-caste interaction increased most in areas with lower caste-based land inequality—precisely the areas where I also find reduced economic reliance on co-ethnics as well as diminished in-group investment. I draw on qualitative data from 56 in-depth interviews to describe the mechanisms through which ethnicity-based social insurance exacerbates ethnic divisions and explain why integration increases only when group inequality is lower. I conducted these interviews in 14 villages with individuals who routinely meet a cross-section of society, such as health workers and local elected officials.
My research explains both the persistence and decline of ethnic divisions in developing countries. Whereas most existing explanations for ethnic divisions emphasize inter-group dynamics of competition and contact, I contribute a framework focusing on within-group economic and social dependencies. Recent scholarship on inter-group relations has built on social psychological theories to devise an array of prejudice reduction tools. My theory and findings complement this literature by showing that there is also an economic content to in-group ties that influences inter-group relations. To the debate on whether welfare undermines community or fosters generalized trust, I contribute the first study from the Global South of how state-provided welfare promotes out-group ties while reducing investment in in-group ties.
This dissertation is organized in nine chapters. Following an introduction in Chapter 1, Chapters 2 and 3 present my theoretical framework. Chapter 2 develops the social investment theory of ethnicity, while Chapter 3 advances my hypotheses about the effects of state welfare provision on inter-group relations, along with the underlying assumptions and scope conditions. Chapter 4 offers a historical perspective on caste, proposing that the contemporary rigidity of caste boundaries may be understood, in part, as a response to the uncertainty and political upheaval of the eighteenth and early-nineteenth centuries.
Chapter 5 examines the functioning of caste-based social insurance arrangements in India, using secondary quantitative data and primary qualitative data. Chapter 6 outlines the political backdrop against which India's first income support program for farmers was launched in the state of Telangana, and situates the state’s welfare spending within the broader Indian context by comparing its social sector and development expenditures to those of Andhra Pradesh, Kerala, and Tamil Nadu.
Chapters 7 and 8 weave together evidence from three data sources—panel data on household loans, an original survey of 3,020 households, and 56 qualitative interviews—to demonstrate how welfare substitutes for caste-based social insurance and enhances inter-caste integration. Chapter 9 concludes by discussing the implications of my findings for electoral politics and political behavior, and shows how key elements of the theoretical framework contribute to our understanding of inter-group relations beyond the Indian case.Political Economy and Governmen
Mechanisms linking pre-mRNA splicing to longevity in C. elegans
No full textGeroscience aims to target the fundamental biology of aging to address the increasing prevalence of age-related diseases. However, a comprehensive understanding of molecular mechanisms underlying longevity interventions and inter-individual variability remains elusive. Prior work in C. elegans determined a critical role of splicing factors SFA-1 and REPO-1 as pathway-specific modulators of longevity, and postulated lipid metabolism, particularly through POD-2, as a downstream effector mechanism. My PhD thesis characterized pre-mRNA splicing and lipidomic changes associated with REPO-1 loss, and determined that POD-2 phenocopies REPO-1 in the regulation of lipid deposits and lifespan extension by dietary restriction, reduced TORC1 and mutant ETC. This work also defined the spatial and temporal effects of REPO-1 loss. These data elucidate molecular mechanisms of responses to longevity interventions. Additionally, my dissertation work demonstrated that differences in Oleic acid concentration underly variation in lifespan expectancy using the ret-1 splicing reporter. This work exemplifies the utility of multi-staged assessment of lifespan variation, thereby significantly contributing to the advancement of precision Geroscience. Altogether, my dissertation substantiates the role of pre-mRNA splicing in both the regulation of response to longevity interventions and the prediction of lifespan expectancy.Biological Sciences in Public Healt
Applications of Computer Vision on the Biology of the Inner Ear
No full textThe hair cells of the sensory epithelium of the peripheral auditory system are precisely tuned for separating frequency, power, and phase from acoustic signals. Enabled by modern imaging techniques, these cells are imaged over large spatial areas and commonly analyzed by hand. Manual analysis is the gold standard in terms of accuracy however, as datasets increase in size the time and labor costs become significant. With careful application, deep learning techniques have demonstrated the ability to capture high level of biological complexity and can aid in automating image analysis. These techniques have potential to save considerable time and drastically expand the amount of imaging data able to be analyzed, leading to less biased research outcomes and new insights. With this work, I have taken considerable steps in maturing the use of deep learning for image analysis of the biology of the inner ear, creating applications and tools across imaging modalities for varied analysis tasks. I have compiled the first open dataset for detection and classification of inner ear hair cells imaged with light microscopy. Using this dataset, I trained the first generalizable deep learning model for hair cell detection and demonstrated its accuracy and usefulness. I have created and validated a novel approach at three-dimensional instance segmentation, demonstrating its usefulness in assessing outer hair cell mitochondria health in normal and pathologic conditions. Finally, I have developed a framework easing the future development and validation of machine learning tools. With these works, I have created a foundation for deep learning image analysis in the auditory field, demonstrated best principles for training, evaluating, and validating deep learning performance, and have created new approaches which have the potential to impact all fields of biology.Medical Science
Human Leukocyte Antigen-E-restricted Killer Immunoglobulin-like Receptor-expressing CD8+ Regulatory T cells in Allograft Rejection
No full textT lymphocytes from CD8 lineage that express natural killer cell receptors have been identified to have regulatory functions. In the context of transplantation, previous studies from our laboratory have shown that murine alloreactive CD4+ T cells upregulate Qa-1 (HLA-E in humans) stress peptide complex on their surface, becoming targets for Qa-1-restricted Ly49+ CD8+ regulatory T cells. Immunization of the host with stress peptides (FL9 and Hsp60) mobilized Ly49+ CD8+ regulatory T cell-mediated killing of alloreactive CD4+ T cells in vivo and improved graft survival in a murine model by inhibiting donor-specific antibody responses. Moreover, inducing a single mutation in Qa1 that prevents TCR binding led to allograft rejection in mice. Preservation of self-tolerance by CD8+ regulatory T cell-mediated pathogenic target cell suppression via self-peptide-MHC1b complex interaction inhibits alloimmune rejection without causing generalized immune suppression.
Translation of murine findings to humans identified inhibitory KIR receptors in humans to be homologous to murine Ly49 in CD8+ regulatory T cells. Our characterization of these KIR+ CD8+ regulatory T cell subsets (KIR+ CD8+ Tregs) in transplantation revealed significantly higher frequency of KIR+ CD8+ Tregs in peripheral blood following antibody-mediated allograft rejection compared to healthy controls. KIR2DL2/3 and KIR3DL1 markers were highly enriched in CD8+ T cells during allograft rejection (median >1%) relative to other inhibitory KIR markers. Inhibitory KIR+ (KIR2DL2/3 and KIR3DL1) subsets in CD8+ T cells were significantly higher following transplantation when compared to healthy controls.
We also demonstrated that KIR+ CD8+ Tregs isolated from human peripheral blood mononuclear cells (PBMCs) can be expanded in vitro using IL-15 and Hsp60 peptide in an antigen-dependent manner. The expanded KIR+ CD8+ Tregs suppressed activated HLA-E-upregulated CD4+ T cells. Additionally, to study human alloimmunity, we developed a humanized mouse model transplanted with human kidney organoid. Using this pre-clinical model, we demonstrated that ex vivo–expanded KIR+ CD8+ Tregs killed alloreactive CD4+ T cells.Graduate Educatio