21793 research outputs found
Sort by
Hybrid Model for Claim Frequency and Claim Severity
Full text linkRate making in insurance refers to the pricing of insurance premiums through calculations, by actuaries, and adjustments in various factors. Fair pricing of insurance products is of utmost importance for insurance companies to be able to face market competition and stay in business. Therefore poor rate making, which could be the result of poor prediction of risks, would be dangerous for insurers.
Insurance data is characterized by an imbalance between the number of policyholders that claim and those that do not. The majority of premium payers do not incur accidents and thus do not claim their losses, resulting in a large number of “zero–claims". However, it is very important for the company to identify the customers who are more likely in future to file a claim, because every claim incurs a cost to the enterprise. Chapter 1 proposes a sampling technique devised to improve the identification of the possible future losses by better tracking of the non–zero claims.
Generalized Linear Models have long been used by actuaries to accomplish the rate making task. The method is parametric and is based on certain assumptions about the distribution of the data. Insurance data with its probability mass at zero, do not fall exactly into the framework of GLMs. However, in the recent years, various new machine learning algorithms have provided improvement by being more effective predictors than GLMs. What these algorithms lack is interpretability. Chapter 2 uses simple algorithms like regression trees, in combination with GLMs, to create a pre–processed GLM that is more effective than a standalone GLM.
The endeavour of improving the classical GLM continues in Chapter 3. Here the another combination of trees and regularized GLMNet is used to produce results with more predictive capability that any one of these algorithms as stand-alone. The new results are interpretable as well as improved
Experimental Evaluation of Rover Mobility Prediction Methods using Single Wheel Experiments
Full text linkMultiple lunar exploration activities from around the world are driving the development of small skid-steer rovers for Moon missions. Single wheel experiments are commonly done to approximate the mobility of a rover design. This research aims to evaluate the limitations and accuracy of simple approximation models. In this work, single-wheel and full rover laboratory experimentation methods are described for systematic comparison of wheel designs in point turns and slope climbing. The effects of lunar gravity are accounted for using Granular Scaling Laws and GRC-1 lunar simulant in all experiments. Single wheel experiments are paired with simple theoretical models to predict full rover performance. The accuracy of the predictions is confirmed using scaled tests with a full rover. Slanted grousers are shown to be compared favorably against V-offset grousers for point turns, achieving successful point turning with approximately 30\% slip in single-wheel and full rover experiments. This also showed that single wheel experiments can accurately predict full rover performance in point turns. On the other hand, slope climbing predictions had a substantial amount of errors. In part, this error was found to be explained by the diameter/width value of the wheel, with larger values being associated to a higher prediction error. To reduce prediction error a correction factor can be approximated by applying a correction factor to rigid wheel test data
Behind the Awakening: Tracing the Visual and Affective Reverberations of the 2021 Colombian National Strike
Full text linkThis thesis looks at the online and offline affective reverberations of the 2021 Colombian National Strike through its visual culture to understand what counter public digital expressions reveal about how Colombians engaged with the events and the affective histories of the nation surrounding these moments. My goal is to show how the interplay between material and digital spaces contributed to the emergence of a new affective structure. I argue that by looking at the images and affects that were shared online, it is possible to see a shift from an affective structure of fear and apathy—promoted by the visuality of the Colombian armed conflict and anti-insurgent propaganda—to a new, and perhaps temporary, structure of “awakening.” By following particular nodes in the strike’s visual and digital affective fabrics, I show how spaces like Twitter (now X) not only offered an affective and political arena to share, discuss and enact practices of resistance, but also became “affective archives” or “counter archives” through which these changes could be traced.
My thesis looks at affect through and as digital images to understand their textures and the ways in which visual activism and digital images touch us. Focusing on two particular moments/spaces that took place during the strike, I use the figure of the hero to understand how these affective changes happened, and, at the same time, how national identity was renegotiated
Aeroelasticity and gust response of wings equipped with a distributed propulsion system
No full textThis thesis studies the aeroelastic behaviour of wings designed for urban air mobility (UAM) through sensitivity analyses, parametric studies, flutter envelopes, and gust responses. As urban traffic congestion intensifies, innovative transportation solutions such as UAM vehicles are gaining attention for their accessibility, reliability, travel time, and reduced environmental impacts. However, their lightweight structures and the versatility of their distributed electric propulsion (DEP) system make them somewhat different from conventional aircraft, thus necessitating fresh investigations into the aeroelastic stability and response of UAM vehicles.
The study employs a wing model based on Euler-Bernoulli beam theory, with propulsors represented as follower forces and an unsteady aerodynamic model in the time domain. The governing equations are discretized using Galerkin's method, with the flutter speed and gust response determined via eigenvalue analysis and time integration, respectively. A building-block approach is employed, starting with a clean-wing configuration, followed by single- then multi-propulsor configurations with tiltrotor capabilities, and finalizing with the implementation of the finite aspect ratio effects.
Sensitivity and parametric studies assess the influence of key wing and propulsor parameters on flutter speed. It is found that some parameters significantly reduce flutter speed, with certain cases leading to a 50 % drop or even preventing flight. It is also observed that distributing the propulsion system across the wingspan is beneficial for the stability of the wing, up to a certain value of thrust. Gust response analyses under sharp-edged and 1-cosine profiles reveal that propulsors strongly influence transient and steady-state deformations. Finite aspect ratio effects increase flutter speed values, highlighting the conservatism of strip theory, but also reduce damping, making gust responses worse in some cases. These findings provide design insights for DEP-equipped wings, guiding future UAM aircraft development. Further work will incorporate propeller slipstream effects and configurations optimization strategies
Deprivatizing Religion in Our Secular Age: Ignatian Pedagogy as a Critique of Laїcité
Full text linkIs there room for religion in the public square? This simple, yet heavy question has been constant and evolving since the Enlightenment period. Prior to this period in the context of Latin
Christendom, the answer was an obvious ‘yes’. The innovative idea that the answer might be ‘no’ has had an immense impact on Church-State relations in various Western contexts in the past 250 years. The experiments in secularity within the past few hundred years wrestling with this question are explored in brief detail. One of the social imaginaries that emerged in France is the concept of laïcité. It is important to address the historical roots of the ideology of laïcité within the broader context of emerging Western secularities. Grievances present within this ideology are also addressed, culminating with a deconstruction of its totalizing features that seek to marginalize other burgeoning ideas of secularity. Ignatian Pedagogy will be illustrated as one of the many examples that illustrate the deprivatization of religion in the public sphere. Furthermore, it will be discussed in terms of its applicability to all students in hopes of both greater human flourishing, and an accomplishment of the possibility for religious freedom and peace among differences, meeting a central goal of the Peace of Westphalia
Irish Big House Basements: A Study of Floor Plans, 1730s–1925
Full text linkExaminations of the basements of Irish country houses are underrepresented in current scholarship, and few historians use as sources the architectural drawings that depict them. This thesis aims to address the resulting gaps by using a sample of 43 floor plans for the basement levels of 29 country houses across the island of Ireland to analyze the work of architects and how these sources can be used to study these buildings that have fascinated the general public for centuries. The plans are examined based on their materiality and as visual layouts, and the considerations facing architects when designing country houses are discussed.
New statistics are produced about aspects of Irish country house basements such as their depth relative to ground-level, how many rooms they contained, and their general functions. Despite its scope remaining within the realm of architectural planning and based solely on floor plans, this thesis generates research useful to architectural and social historians, especially those conducting “history from below,” and suggests several avenues of further research
Nature Reimagined: Christi Belcourt’s Ecological Aesthetic and Its Impact on Canadian Perspectives
Full text linkThis thesis examines how Christi Belcourt’s works, Wisdom of the Universe and Offerings to Save the World, unsettle the legacy of the Canadian landscape tradition by centering the perspectives of Indigenous women and being grounded in traditional knowledge and practices. By situating Belcourt’s works in relation to established landscape paintings by the Group of Seven and Tom Thomson, I demonstrate how her practice challenges romanticized and extractive representations of the landscape, instead highlighting the interconnection of all living beings and the fragility of our ecosystems. Drawing on the work of scholars from Indigenous studies, ecology, Indigenous feminisms, and art history, this thesis examines how Belcourt innovatively adapts storytelling, Indigenous ecologies, and Métis floral beadwork to the canvas to redefine the possibilities of contemporary Indigenous art
An Annotated Translation of the Imperially Commissioned Codex of Manchu Veneration and Thanksgiving
Full text linkThe Imperially Commissioned Codex of Manchu Veneration and Thanksgiving (hereafter Codex), completed in the 12th year of Emperor Qianlong’s reign (1747), is the first independent written record of Manchu Shamanism. It meticulously documents the procedures, hymns, and ritual instruments of a religious tradition practiced in the Qing court, which Qianlong identified as indigenous to the Manchus.
This thesis presents the first annotated English translation of the Codex’s ceremonial programs and hymns. The translations of the programs are based on both the Manchu text and its official Chinese counterpart, while the hymns are translated solely from the Manchu original. Besides providing background on the structure and operations of the Qing imperial court, my annotations explicate the meanings of the Manchu terms for the ritual actions, instruments, and ceremonies, many of which were obscured in the Chinese translation.
In the introduction, I examine the existing translations and research on the Codex, the text’s production, and the two types of rituals mentioned in its title—Veneration and Thanksgiving. I seek to address the gaps in previous scholarships stemming from inadequate attention to the Manchu text and misconceptions on Shamanism. This project lays the groundwork for further research into the Qing court’s Shamanic traditions and their broader cultural, social, and political implications
A TASKS-driven analysis on procrastination behavior
No full textThis thesis aims to investigate the primary factors contributing to procrastination behavior. This exploration is driven by the TASKS framework, a methodology that systematically analyzes human behavior based on affective states, thinking skills, knowledge levels, and available resources. Given the complexity and variability of procrastination behavior among individuals, this study uses two distinct data types to capture the diversity and discrepancies in behavior patterns, ensuring a more comprehensive exploration of procrastination factors. The TASKS framework facilitates data analysis by coding, structuring, comparing, and synthesizing information. Subsequently, a procrastination mechanism is derived from these comparative analysis results. The results indicate that the affective component plays the most significant role in influencing procrastination, with negative feelings exerting the strongest impact. This research also identifies the other potential factors shaping procrastination behavior, with skills and available resources having the second greatest influence, whereas knowledge has the least. This study not only systematically identifies both potential and decisive factors of procrastination, but also assesses the validity of the TASKS framework in explaining its underlying mechanism
Enhancing Hedging Strategies with Deep Reinforcement Learning and Implied Volatility Surfaces
Full text linkThis thesis explores the use of deep reinforcement learning (DRL) to enhance dynamic option hedging by incorporating forward-looking market information, mitigating speculation, and optimizing portfolio rebalancing frequency. The first paper, Enhancing Deep Hedging of Options with Implied Volatility Surface Feedback Information, introduces a DRL-based hedging framework that leverages implied volatility surface data, improving hedging performance over traditional methods. The second paper, Is the Difference between Deep Hedging and Delta Hedging a Statistical Arbitrage?, examines whether deep hedging introduces speculative behavior in incomplete markets, demonstrating that proper risk measure selection prevents unwanted speculation. The third paper, Implied-Volatility-Surface-Informed Deep Hedging with Options, extends deep hedging by integrating implied volatility surface-informed decisions, no-trade regions, and multiple hedging instruments, improving cost efficiency and adaptability. This research contributes by defining frameworks that enhance existing techniques for managing risk in financial markets