18693 research outputs found
Sort by
AI-Driven Optimization of Wind Energy Distribution in Texas Using Multi-Agent Reinforcement Learning
No full textAbstract. The integration of large-scale wind power into modern electrical grids presents persistent challenges due to variability, curtailment, and compliance with operational constraints. This study proposes a multi-agent reinforcement learning (MARL) framework for optimizing wind energy distribution within the Texas power grid. The system employs three specialized agents—managing wind curtailment, storage utilization, and load adjustments—to collaboratively balance supply and demand under dynamic grid conditions. Using historical operational data from the Electric Reliability Council of Texas (ERCOT), the framework was trained and evaluated on a range of scenarios encompassing both typical and extreme operating conditions. Results demonstrate substantial performance improvements compared to baseline dispatch strategies, including a measurable reduction in supply–demand mismatch, improved storage state-of-charge stability, and enhanced coordination among agents. The approach offers a scalable, adaptable, and regulation-compliant pathway for renewable integration in grids with high penetration of variable energy resources
Texas, Delaware, and the New Controller Primacy
No full textIn January 2024, Elon Musk, CEO of Tesla, Inc., initiated the process of moving Tesla’s state of incorporation from Delaware to Texas, citing concerns over Delaware corporate law. The move sparked discussions about the dominance of Delaware corporate law and its recent rulings, with critics accusing Court of Chancery judges of favoring shareholder interests over controller innovation and managerial discretion.
The shift away from Delaware is set against a backdrop of increasing judicial scrutiny in the Delaware Court of Chancery, where plaintiffs have recently seen successes in litigation against controlling shareholders, corporate directors, and officers. Concerns over Delaware’s evolving legal landscape have prompted other corporations led by controlling shareholders, including TripAdvisor, Inc., The Trade Desk, Inc., Meta Platforms, Inc., and Trump Media & Technology Group, to consider reincorporation elsewhere. Despite Delaware’s long-standing status as the preferred state for corporate governance due to its well-established case law and judicial expertise, recent legislative battles and judicial decisions have introduced the possibility of more scrutiny for controller transactions.
Texas, while historically not a hub for corporate incorporations, has emerged as a potential alternative due to its business-friendly environment and the recent establishment of specialized business courts. However, Texas lacks the robust body of corporate case law that Delaware offers. The move by Tesla and the broader conversation around Texas corporate law suggest an emerging shift where businesses may seek to shape corporate governance frameworks more favorable to managerial control, potentially establishing a new “controller primacy” model. Legislative developments in early 2025 in both Texas and Delaware signal an ongoing competition between the two states to attract corporate incorporations, raising questions about whether they are racing to the top or the bottom in corporate governance. However, the Texas legislature’s move in May 2025 to allow listed corporations to bar derivative suits from almost all shareholders, an opportunity that the Tesla board immediately seized upon, seems to set a new bar in defending against shareholder suits
Procedural Generated World With History Simulation and 3D Map
No full textThe procedurally generated world with history simulation tool dynamically generates customizable, seed-driven virtual worlds through algorithmic processes. Using Voronoi diagrams (constructed via Fortune’s algorithm) for spatial partitioning, it creates diverse geographic features such as lands, oceans, and heights. Climate systems, hydrology (rivers, precipitation), and temperature data are procedurally derived from the terrain data. The tool further simulates sociopolitical dynamics, including population distribution, cultural evolution, religious spread, and diplomatic interactions between procedurally generated nations. Cities, towns, roads, and products emerge organically based on province data.
Users can interactively modify generated worlds—adjusting province geographic data, reshaping cultural or religious distribution, or changing province owner—before exporting structured data (saved as XML) for compatibility with external applications
The Champion of Images: A NeuroAI Framework for Understanding Image Effects on Consumer Decision-Making
No full textIn today\u27s e-commerce landscape, consumers\u27 first interactions with products often occur through product thumbnails, called champion images . Champion images are common in search results and influence click-through behavior. This research investigates what makes these images click-worthy by developing a novel NeuroAI approach-an image mining method that bridges neuroscience and computer vision models. Using a large-scale open-source fMRI dataset, we trained prediction models of neural responses to images and conducted two novel fMRI studies to validate this approach. We applied these models to analyze a global online travel agency platform\u27s hotel search clickstream data. Examining the modeled neural responses to champion images in hotel listings, we found that images with lower neural processing demands were more likely to attract click-throughs, which persisted after controlling for hotel quality, price, and image type. We identified three key neurally-informed features that could influence click-through behavior: retinotopic demand, implied motion, and navigational affordance. Our approach explains additional variance in click-through behavior beyond existing non-neurally informed image metrics. This research establishes a neurobiologically grounded framework for understanding visual marketing, offering theoretical insights into processing fluency in online search environments and practical recommendations for optimizing product images to improve click-through rates
An Application of the Archaeology of the Human Experience to Classic Period Hohokam Burials at S\u27edav Va\u27aki, Phoenix, Arizona
No full textHohokam archaeologists have long viewed the Classic Period (1100-1375 CE) in central Arizona as a time of rapid and profound change (Doyel 1981:55; Abbott 2003b). After the Pre-Classic Period (450-1100 CE), populations coalesced into massive villages each containing one or more platform mounds (Abbott et al. 2021; Bayman 1996); Hohokam farmers built and maintained one of the largest networks of canals in the Americas, and socio-political complexity reached its peak (Abbott 2003b; Abbott et al. 2006; Haury 1976). However, prosperity was tempered by possible overpopulation, drought, floods, warfare, and social change, especially in the Late Classic (Abbott 2003b; Benson and Berry 2009; Greenleaf 1975; Huckleberry et al. 2018).
Until relatively recently, most archaeologists have relied on interpretive assumptions that portray a catastrophic end to the Hohokam (Ezell 1961, 1963; Fish and Fish 2007), despite increasing evidence and widespread recognition of Hohokam cultural continuity (Loendorf and Lewis 2017). Current reconstructions offer new perspectives regarding the relationship of descendant communities, particularly the Akimel O’Odham, with Hohokam ancestors that seek to examine social change in ways that do not rely on preconceived notions of socio-political organization and social evolution (Borck and Clark 2023; Loendorf and Lewis 2017). These new approaches emphasize social change, demography, and the organizational shifts that lead to current and historical expressions of the material culture and archaeology of descendent communities. The transitional period between the late Classic Hohokam and the Post-Classic is particularly salient to these new perspectives. It was during this period that platform mounds ceased to be used as ceremonial residences and the center of population shifted south from the lower Salt River to the middle Gila River (Abbott 2002, 2003b; Doelle and Wallace 1991; Wilcox 1989).
The goal of this dissertation is to understand the factors that affected security and well-being among the Classic Period Hohokam and by examining the nature of demographic and social inequalities that emerged during the reorganization of the Hohokam cultural system. This research draws from the Archaeology of the Human Experience (AHE) to explore how people coped with rapid or widespread change and how it affected their lives (Hegmon 2016). The analysis uses field records to examine mortuary features from S’edav Va’aki (AZ U:9:1(ASM)), in Phoenix, Arizona, which was analyzed to understand several dimensions of human security including economic, food, health, and personal/community security. Economic, food, health, and personal/community security during the Hohokam Classic Period were examined using mortuary accompaniments, burial architecture, and osteological pathologies in the S’edav Va’aki dataset.
Statistical analysis of these variables revealed that in the Early Classic, there were significant demographic differences in mortuary accompaniment richness, abundance, and distribution. Adults, overwhelmingly males, had more items than subadults and females. However, this trend wanes in the Late Classic, with all demographics becoming more equal in mortuary richness, abundance, and distribution. There are no significant differences in health diachronically by sex or between the sexes in either period. Increased labor investment in burial architecture is evident over time. Based on the results of this demographic analysis, only economic security declined diachronically, primarily for adults and males.
My analyses also show that the proportion of individuals buried with an abundance of mortuary accompaniments declined over time. This suggests that rare and/or important mortuary accompaniments were concentrated in the hands of fewer individuals, including subadults, over time. The data suggest that economic security, as measured by the access to rare or important items, declined. This could indicate a reduction in trade and/or a change in social systems affected those at the top more than the general population as documented by others (see Abbott 2003a, 2003b). These conclusions mirror O’Odham oral tradition which details an accumulation of power by platform mound leaders in the Late Classic, their loss of power, and reorganization in the Post-Classic (Borck and Clark 2023; Bahr 2001).
Ultimately, this research demonstrates that human security changed over the Hohokam Classic Period in ways that have been described by descendant communities. Health security and personal/community security as measured by the variable discussed here appears to be stable through time, while economic security declined in the adult male subset of the population. This study showcases the utility of legacy collections in new archaeological research, and it also illustrates the insights that different aspects of mortuary archaeology (mortuary accompaniments, burial architecture, and health of individuals) can provide in understanding how social and environmental change affected past peoples. Additionally, this research shows the value of collaborating with descendant populations and inclusion of traditional knowledge
Hybrid Graph-Recurrent Architecture for Citation Recommendation via Future Embedding Forecasting
No full textThe rapid expansion of scientific literature has intensified the challenge of identifying relevant citations, particularly for newly published or under-cited papers. Traditional citation recommendation systems typically model static relationships or respond to past citation activity, offering limited predictive power for emerging works. In response, this thesis presents a temporal modeling framework for citation recommendation that anticipates future scholarly relevance by forecasting the latent representations of academic papers.
Building on prior work that utilized Temporal Graph Networks (TGNs) to model dynamic citation flows, we propose Graph-Time, a hybrid architecture that integrates a Graph Transformer with a GRU-based time series predictor. The transformer captures structural and temporal dependencies within yearly citation snapshots, while the GRU forecasts a paper’s future embedding based on its past trajectory. To address cold-start cases, We inc
We evaluate Graph-Time on the AMiner Citation Network V14, comprising over 5 million papers and 36 million citation links, using strict year-based splits to avoid information leakage. Compared to dynamic and static baselines, our model achieves substantial improvements in recommendation performance, including a 21 percent gain in MRR and notable gains in Precision@10 and Recall@10. Ablation studies confirm the contribution of each architectural component, and we discuss ethical considerations surrounding bias reinforcement in scholarly recommendations.
This work contributes a scalable, future-aware framework for citation forecasting and lays a foundation for real-time, temporally sensitive academic recommender systems
Hyper Modularity and Dynamic Materials in the Style of Thomas Kinkade
No full textMy thesis is a small treehouse village, built using hyper modularity and designed to style-match Thomas Kinkade’s art. Dynamic materials are integrated into the scene as part of style match. The village, human-sized, is constructed on a massive ancient tree. Magical elements are incorporated throughout the village, brought to life through dynamic materials
Towards Reliable Clinical Applications of AI Models in Radiotherapy
No full textOver the past decade, artificial intelligence (AI), particularly through deep learning (DL) techniques, has made significant strides in fields like computer vision (CV) and natural language processing (NLP), leading to transformative advancements across numerous applications. This progress has sparked considerable enthusiasm within the medical field, where DL-related research has grown exponentially since 2015. However, despite these promising developments, the real-world deployment of DL models in healthcare remains limited, especially in safety-critical domains such as radiotherapy (RT), where reliability, safety, and sustained performance are critical. This thesis addresses three core challenges associated with the clinical application of DL models: (1) post-deployment performance degradation, (2) the lack of reliable, case-specific quality assessment for DL predictions, and (3) the absence of robust, generalizable performance monitoring frameworks tailored to dynamic clinical environments.
First, we examine the long-term performance patterns of DL models deployed in clinical settings, focusing on their effectiveness in adapting to evolving clinical practices. Through retrospective simulation using prostate cancer RT data from 2006 to 2022, we demonstrate a notable decline in auto-segmentation performance over time, attributed to changes in clinical practices, personnel shifts, and the introduction of new techniques. These findings underscore the necessity of continuous evaluation of model performance beyond initial validation.
Second, we propose a novel auto-contour quality assessment (QA) framework for DL-based segmentation in RT, specifically designed for online adaptive radiotherapy (OART). Our approach integrates Bayesian ordinal classification with uncertainty quantification to provide case-specific, uncertainty-aware quality assessments, accommodating various scenarios with limited or no manual labels. By incorporating an additional calibration step, our method can achieve clinical accuracy exceeding 90% as required for confident predictions. The proposed AI-assisted auto-contour QA model effectively streamlines contouring processes, substantially reducing manual effort and improving clinical workflow efficiency in OART. By integrating uncertainty quantification, our approach enables clinicians to make rapid, informed decisions, ensuring improved patient safety and workflow reliability in time-sensitive clinical workflows.
Third, we introduce DyMon, a dynamic monitoring framework utilizing empirical prediction interval coverage rates (EPCR) derived from conformal prediction, combined with adaptive statistical testing methods, to continuously monitor deployed AI models. EPCR serves as a robust, model-independent indicator, identifying distribution shifts by detecting deviations from predetermined coverage levels. We assess DyMon using real-world auto-segmentation data, validating its performance through three adaptive statistical tests: Bayesian adaptive testing, Window-limited Generalized Likelihood Ratio Cumulative Sum (WinLCUSUM), and Maximized CUSUM (MaxCUSUM), each suited to different change patterns. Our results demonstrate that DyMon effectively identifies performance deterioration in a timely manner while maintaining a controlled Type I error rate
Jurisdictional Competition and Corporate Law: The Rise of Delaware and the Fall of New Jersey
No full textThis paper examines one of the most consequential jurisdictional shifts in American corporate history: the transition of corporate charter dominance from New Jersey to Delaware. During the late nineteenth and early twentieth centuries, New Jersey’s liberal incorporation statutes, including the legalization of holding companies, positioned it as the premier domicile for America’s industrial giants. However, Governor Woodrow Wilson’s progressive reforms in 1913—known as the “Seven Sisters” Acts—reversed that trajectory, imposing significant regulatory constraints and catalyzing an exodus of corporations. Delaware, with its permissive General Corporation Law, expert Court of Chancery, and constitutionally protected legal stability, emerged as the preferred alternative. This article explores the legislative, political, and institutional dynamics that facilitated Delaware’s ascendancy and highlights the enduring implications for federalism and jurisdictional competition in corporate governance. By tracing this pivotal historical realignment, the analysis offers insights into how legal infrastructure, not just statutory text, sustains long-term dominance in the market for corporate charters
Glass Half Full: Strategic Packaging Decisions and Environmental Outcomes
No full textProblem definition: Amid rising environmental concerns and regulatory pressures, firms, particularly beverage manufacturers, face complex trade-offs when choosing among packaging options such as non-recycled plastic, recycled plastic, and glass. While glass is often perceived as the most sustainable option, its true environmental impact depends critically on return and reuse rates, which are influenced by firm-led deposit-refund systems and sustainability advertising. Methodology/results: We develop an analytical model to study a firm\u27s joint decisions on packaging type, pricing, and advertising efforts in a setting with heterogeneous consumer environmental awareness and greenwashing risk. We quantify the conditions under which glass can outperform plastic in terms of profit and emissions---or, paradoxically, lead to higher environmental harm. Under optimal pricing and advertising efforts, the firm can achieve a rare win-win-win : prevent potential greenwashing while simultaneously improving profits and customer utility. We extend the model to examine the effects of recycled content mandates and endogenous deposit values, showing how firms can balance deposit incentives and advertising strategies to engage both environmentally aware and unaware consumers. A calibrated numerical study using data from a regional U.S. dairy firm reveals that adopting the model’s recommended policy—glass packaging with a premium price and targeted advertising—can increase profits by almost 60% while reducing emissions by over 30%. Managerial implications: Our findings can help identify when glass containers outperform plastic alternatives and guide how sustainability advertising and refundable deposits can be leveraged to boost glass return rates, improve environmental outcomes, enhance customer utility, and increase profits