49755 research outputs found
Sort by
Climate-change disputes: is there a role for international arbitration?
No full textThis Perspective argues that international arbitration can and should play a primary role in the resolution of climate-change disputes, rather than playing a secondary role in the shadow of climate litigation, considering both its comparative advantages over litigation and the evolving requirements of EU corporate sustainability due-diligence rules
The Colombia-US 2025 joint interpretation: clarifying investment standards or quietly reshaping investor protections?
No full textThe Colombia-US 2025 joint interpretation (JI) is the first adopted by the United States since the 2001 NAFTA JI. While it favors state regulatory space, this Perspective argues that its restrictive approach to some provisions and application to pending cases raise familiar concerns about the impact and legitimacy of JIs
Production Complexity as Appropriation Protection
No full textScholars have long suggested that a technology gap between firms and developing countries protects firms' assets. How can we capture this intuitive idea that what firms make and how they make it affects their appropriation risk? And under what conditions would we expect firms to rely on a technology gap as a risk-mitigating strategy?
This dissertation aims to address these questions. I make three main arguments: First, firms' production complexity---the products they make and the process by which they make them---affects their relationship with governments. Second, production complexity can protect firms from appropriation by increasing the technical expertise bureaucrats need to appropriate. Third, bureaucratic capacity with respect to firms moderates the relationship between firms' production complexity and their appropriation risk.
This theory leads to two testable hypotheses: (1) bureaucratic capacity moderates the relationship between firms' production complexity and their appropriation risk and (2) in states with low bureaucratic capacity, firms with more complex production will have lower appropriation risk than firms with less complex production. I test these hypotheses by relying on both in-country interviews and firm-level panel data. The qualitative analysis draws on both cross-country variation in bureaucratic capacity and within-country variation in production complexity. The quantitative analysis holds bureaucratic capacity constant by focusing on one country, leveraging variation in firms' production complexity and their experiences with appropriation.
There are three key findings of this research. First, production complexity has a negative relationship with costs of appropriation to firms. Second, a lack of bureaucratic capacity with respect to firms is the causal mechanism driving this negative relationship. Third, production complexity has a positive relationship with bureaucrats' efforts to appropriate.
These findings demonstrate that production complexity is both a firm-level determinant of firms' appropriation risk and something firms can rely on to protect themselves from appropriation in contexts with low bureaucratic capacity. They also suggest that bureaucrats and firms play a ``cat-and-mouse game,'' where production complexity makes firms a target for appropriation, while also decreasing how much is actually appropriated. This dissertation contributes to research on firm-level determinants of appropriation risk; strategies firms use to protect themselves from appropriation; and the ways in which firms' protective strategies affect their political power and behavior, as well as states' capacity
Robot Learning with Sparsity and Scarcity
No full textUnlike in language or vision, one of the fundamental challenges in robot learning is the lack of access to vast data resources. We can further break down the challenge into (1) data sparsity from the angle of data representation and (2) data scarcity from the angle of data quantity. The data sparsity problem means that there is a large proportion of empty space or non-relevant information in the data we collected. Robotics is the science of interaction. We have a piece of software or an algorithm embodied inside a piece of hardware, and then the robot needs to interact actively with the environment to collect useful information. The sequential manner of such interaction and the lapse between two consecutive actions make robotic data inherently very sparse. On the other hand, the data scarcity issue is that the sheer amount of data we can collect in the domain of interest is very limited. In contrast to the richness and accessibility of text, image, and video data available on the Internet, it is extremely difficult to collect data from physical hardware or humans on a large scale.
In this thesis, I will discuss my PhD work on two selected domains: (1) tactile manipulation and (2) rehabilitation robots, which are exemplars of data sparsity and scarcity, respectively. Tactile sensing is an essential modality for robotics, but tactile data are often sparse, and for each interaction with the physical world, tactile sensors can only obtain information about the local area of contact. I will discuss my work on learning vision-free tactile-only exploration and manipulation policies through model-free reinforcement learning to make efficient use of sparse tactile information.
On the other hand, rehabilitation robots are an example of data scarcity to the extreme due to the significant challenge of collecting biosignals from disabled-bodied subjects at scale for training. I will discuss my work in collaboration with the medical school and clinicians on intent inferral for stroke survivors, where a hand orthosis developed in our lab collects a set of biosignals from the patient and uses them to infer the activity that the patient intends to perform, so the orthosis can provide the right type of physical assistance at the right moment. My work develops machine learning algorithms that enable intent inferral with minimal data, including semi-supervised, meta-learning, reciprocal learning, and generative AI methods
Play with the changes: Innate rules for learned vocal communication in songbirds
No full textBirdsong has been a muse of philosophers, artists and scientists for millennia because it is complex, intricate, and shares many features with human speech and music. Like humans and unlike most other animals, songbirds are vocal learners and communicators who use auditory skills developed in early life to shape their adult songs.
Learning confers uniqueness to each individual’s song, but the songs of individuals within a species are structurally similar, particularly in their temporal and sequential organization. The extent to which species identity drives song temporal organization, and how perceptual systems control the auditory encoding of song versus song elements are largely unknown. In this dissertation, I use comparisons of species that differ in song behavior to test hypotheses about the inborn and experiential forces that organize adult song.
Chapter 1 tests the hypothesis that song temporal organization is explained by species rather than learning by analyzing the songs of birds who were tutored, untutored, tutored by heterospecifics, or were hybrids of two species. Chapter 2 uses computational modeling techniques to test the hypothesis that the adult songs of pupils and their tutors are similar because they are the same species. Chapter 3 is an electrophysiology study that tests the hypotheses that a secondary auditory cortical region separately processes song elements versus temporal pattern, and that species-specificity for temporal pattern is a circuit-level property.
Results indicate that learned vocal communication behavior in songbirds is organized by innate rules that are, in part, determined by the species-specific encoding capacities of auditory circuits
Thulium-doped Avalanching Nanoparticles
No full textInnovations in optics and photonics are crucial to propel forward the technologicaladvancement of our society. In particular, the phenomenon of photon upconversion using lanthanide ions holds great potential for many applications such as solar energy harvesting, bioimaging, and anti-counterfeiting technologies. During my journey of studying lanthanide- doped upconverting nanoparticles, I helped discover the first nanoparticles doped with Thulium (Tm) ions that exhibit the photon avalanching (PA) behavior. PA is a special case of photon upconversion; our discovery, called avalanching nanoparticles (ANPs), was exciting for the field because it exhibits highly optically nonlinear behavior and increases the quantum efficiency of the upconversion process by an order of magnitude compared to regular upconversion, thus opening up new applications. In this thesis, three major applications are explored: viral disinfection, photoswitching, and bioimaging.
To study how ANPs can be applied for viral disinfection, we used photoluminescence measurements to observe the UV emission from ANPs when pumped with a 1064nm laser. The emissions at 360nm and 340nm correspond to the 1D2 to 3H6 and 1I6 to 3F4 energy level transitions in Tm respectively. While the 290nm emission, corresponding to the 1I6 to 3H6 transition, was not observed due to limitations within the optical setup, it is safe to infer that they do exist since the 340nm emission comes from the same excited state. All three of the aforementioned emissions are within the UV spectrum and are germicidal. To understand how many ANPs, and how much pump power is needed to disinfect a surface area, we proposed to incorporate ANPs inside N95 masks, and disinfect them with NIR (1064nm) light. We modeled various parameters: laser power (10- 100W), ANP in polymer volume (1.5-15%) and the quantum yield of ANPs (0.3-10%). Most of the calculation yielded a faster disinfection rate than the current state-of-the-art. Disinfecting with NIR light compared to UV is thus not only more efficient, but is also better and safer for the materials and humans present.
While doing measurements on ANPs, I unexpectedly observed them blinking on and off mid-measurement. This is surprising because they were supposed to be extremely photostable according to all the literature on upconverting nanoparticles in the last couple of decades. To better understand this, we did various experiments to rule out any external possibilities with photoluminescence microscopy, atomic force microscopy, and temperature measurements; we concluded the blinking was an intrinsic property of the Tm ions. We learned to deterministically control the blinking process, In the end, we demonstrated several potential applications with the photoswitching property of ANPs, including super resolution imaging.
To demonstrate bioimaging with ANPs, we incubated ANPs in MDA-MB-231 cancer cells rich with receptors. We showed that just like single ANPs on a coverslip, ANPs in cells can also show sub-100nm resolution when imaged with a laser scanning confocal microscope. To better understand how ANPs are distributed within a given cell, 3D sectioning is performed on a cell with 11 frames in the XY direction, 1 um away from each other in the Z direction. With quantitative analysis of the distribution of ANP emission, we concluded that the ANPs were distributed mostly as monomers, with decreasing chance of forming small clusters of dimers, trimers, and quadruplets. This could infer the clustering tendencies of the cancer cell receptors. Imaging with ANPs as the probe thus offer the capability of conducting quantitative analyses that fluorescent probes don’t
In Search of Renewal: Women of Color as Leaders and Change Agents in Higher Music Education
Get PDFRooted in long-standing histories of racial, class, and gender exclusion, Higher Music Education Institutions continue to grapple with acknowledging diversity as a pillar of decolonization. This multiple-bound case study centers the testimonios of three professional Latina singers regarding their experiences in music schooling and four female leaders of color and their work in a highly competitive music conservatory.
Foregrounding institutional change, this study explored how these women of color disrupt narrowly defined processes of knowledge production in higher music education by enacting grassroots, bottom-up policy practice and the oppositional consciousness of Third World Feminism.
More specifically, it examined how these women (1) navigate institutional policy structures, (2) understand the power dynamics shaping legitimacy discourse, and (3) enact agentive strategies. The collaborative data production process revealed that Latina singers resisted cultural erasure and Eurocentric biases by building community, advocating for Latin American music, and challenging exclusionary standards.
The female leaders of color confronted White, male-dominated leadership norms that perpetuated tokenistic diversity and silencing. Their strategies include code-switching, data-driven advocacy, and community-based leadership models to foster meaningful inclusion and disrupt hierarchical power structures. Together, the women in the study advocate for institutional policies that (1) promote meaningful inclusion, (2) redefine success and excellence, (3) prioritize well-being, (4) foster supportive communities, and (5) embrace art as civic engagement
Top-down vs bottom-up processes: A systematic review clarifying roles and patterns of interactions in food system transformation
Get PDFUrgent calls for food system transformation have spurred a variety of responses globally. In some cases, these calls have been answered through top-down led processes, driven by public agencies to design and implement measures that can drive societies towards more viable patterns of development. In other cases, transformation processes have been ignited by community level actors who addressed sustainability issues with context-specific solutions. The broad range of actors raises the question of whether it is top-down or bottom-up processes and actors that are better placed to deliver the fundamental and system level changes that characterise transformation. Through a systematic review, we identified 40 case studies across 24 countries to investigate the role of top-down or bottom-up processes in transformation, whether the two might intertwine, and with what results. We propose five different types of interactions: Autonomous Bottom-Up, Collaborative Bottom-Up, Top-Down Struggles and Resourceful Bottom-Up, Collaborative Top-Down and Transformation Alliances. Based on our analysis, we propose a new heuristic of roles and interactions between different actors. We suggest a shift from dichotomic views on top-down and bottom-up actor roles towards the concept of “transformation functions,” which would re-centre the discussion around the existing or needed capabilities for transformation in different contexts. Finally, we call for further research to determine how different transformation functions need to become more synchronised -or coordinated-to accelerate transformation
Unlocking the link: predicting cardiovascular disease risk with a focus on airflow obstruction using machine learning
Get PDFBackground
Respiratory diseases and Cardiovascular Diseases (CVD) often coexist, with airflow obstruction (AO) severity closely linked to CVD incidence and mortality. As both conditions rise, early identification and intervention in risk populations are crucial. However, current CVD risk models inadequately consider AO as an independent risk factor. Therefore, developing an accurate risk prediction model can help identify and intervene early.
Methods
This study used the National Health and Nutrition Examination Survey (NHANES) III (1988–1994) and NHANES 2007–2012 datasets. Inclusion criteria were participants aged over 40 with complete AO and CVD data; exclusions were those with missing key data. Analysis included 12 variables: age, gender, race, PIR, education, smoking, alcohol, BMI, hyperlipidemia, hypertension, diabetes, and AO. Logistic regression analyzed the association between AO and CVD, with sensitivity and subgroup analyses. Six ML models predicted CVD risk for the general population, using AO as a predictor. RandomizedSearchCV with 5-fold cross-validation was used for hyperparameter optimization. Models were evaluated by AUC, accuracy, precision, recall, F1 score, and Brier score, with the SHapley Additive exPlanations (SHAP) enhancing explainability. A separate ML model was built for the subpopulation with AO, evaluated similarly.
Results
The cross-sectional analysis showed that there was a significant positive correlation between AO occurrence and CVD prevalence, indicating that AO is an important risk factor for CVD (all P < 0.05). For the general population, the XGBoost model was selected as the optimal model for predicting CVD risk (AUC = 0.7508, AP = 0.3186). The top three features in terms of importance were age, hypertension, and PIR. For the subpopulation with airflow obstruction, the XGBoost model was also selected as the optimal model for predicting CVD risk (AUC = 0.6645, AP = 0.3545). SHAP shows that education level has the greatest impact on predicting CVD risk, followed by gender and race.
Conclusion
AO correlates positively with CVD. Age, hypertension, PIR affect CVD risk most in general. For AO patients, education, gender, ethnicity are key CVD risk factors
Cathleen McCarthy
Get PDFIn 1920, a thirty-year-old woman named Cathleen McCarthy got a job as a reporter at The Peterborough Examiner. As the only woman reporter on staff, she quickly found herself in charge of the “Women’s Page” and its “society” and fashion features. Within a year or so, however, she was also writing film reviews—and would continue on that beat until she left the paper in 1937. McCarthy was not only the first film critic in Peterborough, but also one of a very few in Canada