1424 research outputs found
Sort by
Interior-point algorithms for solving linear complementarity problems and their applications
Full text linkIn this thesis, we present the theory of linear complementarity prob-lems (LCPs), with a particular focus on problems whose matrix belongs to the class of P∗(κ)-matrices. In the introduction, we also provide a brief historical background of the theory of interior-point algorithms (IPAs). We discuss the concept of algebraically equivalent transforma-tion technique (AET) and outline the role of this method in the theory of IPAs. To motivate our research, we present several problems that can be written as LCPs. We consider the Arrow-Debreu competitive market equilibrium problem with linear and Leontief utility functions, which can also be formulated as an LCP.
The main body of the thesis focuses on our research results. In Chapter 2, we introduce a new class of AET functions that defines a primal-dual (PD) IPA for P∗(κ)-LCPs. We also prove that the IPA has polynomial iteration complexity in the size of the problem, the parameter κ, the accuracy parameter and the starting point’s duality gap. In Chapter 3, we propose another class of AET functions leading to a predictor-corrector (PC) IPA for solving P∗(κ)-LCPs and prove that it has the same complexity as the currently known best ones for IPAs. Chapter 4 extends our research results by presenting a new PC IPA operating in a wide neighborhood of the central path. We generalize the concept of the wide neighborhood of the central path by using the AET technique. Furthermore, we present the complexity analysis of this IPA, as well. Finally, the thesis includes extensive numerical experiments for both the PD and PC IPAs, along with newly generated test cases to demonstrate the efficiency of the algorithms. The concluding chapter summarizes the obtained results and outlines several promising directions for future research
"The Optimistic Blindspot: Hungarian AI Developers and the Future of Work"
Full text linkKutatásomban az MI és a kapcsolódó technológiák, valamint a munka jövőjéről folytatott vita kapcsolatát egyedülálló szemszögből vizsgálom, a legújabb MI-fejlesztésekben jártas magyar IT-szakemberek szemén keresztül. Ennek több oka is van. A munka jövőjéről szóló irodalmat széles körű szakértők írták, akik mind saját szakterületük különböző nézőpontjait hozzák be. A központi kérdés a legtöbb esetben az, hogy vajon a robotok elveszik-e a munkánkat. Az általuk képviselt állásponttól függően a szerzők vagy azt állítják, hogy egy átalakító és radikális változás küszöbén állunk, amely drasztikusan csökkenti az emberek arányát a munkaerőpiacon, előre nem látható következményekkel járva (Frey & Osborne, 2013; Ford, 2017; Brynjolfsson & McAfee, 2016; Huws, 2014), vagy ellenkezőleg, az ilyen állításokkal szemben érvelnek, és egy lassú, kezelhető átmenet mellett szólnak (OECD, 2016; Kelly, 2023). Ritkán választják a harmadik utat, és sürgetik egy teljesen gépi alapú gazdaság eljövetelét, amely megszabadítaná az emberiséget a gazdaságilag kényszerített munka terhétől (Bastani, 2019). Az irodalom nagy része általános kijelentésekre összpontosít, amelyeket a globális észak akadémikusai írnak, korlátozott példákkal Afrikából (Chigbu & Nekhweva, 2021), Mexikóból (Lovett et al., 2004) vagy Indiából (Bisht et al., 2023; Lingmont & Alexiou, 2020). Az Európai Bizottság adatai széles témákkal foglalkoznak, és olyan adatokat szolgáltatnak, amelyek segítenek Magyarországot Európában elhelyezni. Az adatok makroszintű jellege azonban nem teszi lehetővé, hogy részletes állításokat tegyünk a részt vevő országokról (European Commission, 2020, 2021). A magyar kutatók alapvető munkái (Tardos & Ságávri, 2021; Keszey & Tóth, 2020; Makó & Illéssy, 2020; Illéssy et al., 2021; Fehér & Veres, 2022) ezért nagy segítséget jelentettek a magyar kontextus többrétegű képének felvázolásában az MI és az automatizálás jövőbeli hatásai kapcsán. A magyarországi kutatások révén megismerhetjük a szakemberek széles körének véleményét és elvárásait, a kék galléros gyári munkásoktól a mérnökökön és vezetőkön át a szakszervezeti képviselőkig (Tardos & Ságávri, 2021; Keszey & Tóth, 2020) és a Magyar MI Koalíció tagjaiig (Fehér & Veres, 2022), míg a magyar kormány MI stratégiájáról szóló hivatalos jelentés a politikai szféra elvárásait tükrözi (Magyarország Mesterséges Intelligencia Stratégiája 2020-2030, 2020).
Ez a PhD disszertáció az OTKA K131733 számú, "A mesterséges intelligencia és a társadalom víziói" című kutatás része, és így része egy sor tudományos vizsgálatnak, amelyek különböző módszertanokat alkalmaznak és különböző társadalmi csoportokat céloznak meg. A kutatások során a fiatal magyar felnőttek elvárásait vizsgálták (Vicsek et al., 2022). Az MI szakértők részt vettek egy backcasting workshopon, amely az optimális jövővel kapcsolatos elvárásaikat hasznosította (Pintér, 2023), míg ez a kutatás scenario-building módszert alkalmazott. A csapattagok segítettek a disszertációban alkalmazott módszertan kidolgozásában
The Impact of Oil Production and Consumption on the Balance of Power
Full text link- My research examines how states can exercise power through oil. What kind of balancing strategies they use to maintain or improve their power. I use quantitative data and the case studies of USA, Japan and USSR/Russia to research the question.
- The novelty of my research is enhanced by the international relations theory it applies to try to come up with answers: Neoclassical realist theory.
- The neoclassical realist approach incorporates both systemic stimuli (such as neorealism) when analysing a policy response, but it incorporates other variables such as leader perception, strategic culture, etc. as well in order to have a better explanatory power.
- The case of oil and balance of power is more optimal if it involves unit level variables into the research as well, because the systemic forces cannot solely explain the policy responses of states and their international outcomes. As all realist theories, neoclassical realism also considers international relations and politics as a struggle for influence and power where the resources are finite in a world of uncertainty about each state’s intentions and capabilities.
- The macro-periods that I divided my research are distinct from each other by their characteristics: 1918-1945: The period between the end of the First World War and the end of the Second World War. This timeframe is very much different from the post 1945 world and shows the early conquest of oil. 1945-1973: After the Second World War the balance of power shifted tremendously and the era of the Cold War deserves its own macro-period in my research. This is the era of cheap oil. 1973-1998: The end of cheap oil, the fall of the Soviet Union and the Bipolar World signals a new chapter in the international relations. Furthermore, this is the era of the decline of conventional oil sources in the USA. 1998-Present: The post 9/11 world and the increasing relevance of non-state actors in the international relations. The Middle East, a significant oil-producing region moves to the forefront of attention. The introduction of shale oil.
- H1. Every Great Power had access to the majority of the proved oil reserves since the Industrial Revolution. In the case of H1. I aggregated the available data in Table 19. to show the distribution of reserves in the proposed macro periods. In the case of H1. I found enough evidence to conclude that this hypothesis is correct.
- H2. Ensuring access to crude oil influences the foreign policy of Great Powers. The less oil resource one state has control over, the more its foreign policy is focused on it. In the macro periods, historical examples and with the case studies I did not find enough supporting evidence for H2. Furthermore, in the case of the United States and Russia I found several contradictory actions that show: even oil rich states focus their foreign policy around the access to oil. H2. can only be correct when a Great Power has control over abundant oil sources, there is a short period of disruption of oil supply (or the crisis does not impact the global oil production) furthermore, it has enough surplus reserves that can boost production and there are direct or indirect means of increasing daily oil production independent of the crisis.
- H3. Great Powers try to maintain and/or improve their place in the international system through balancing strategies. When Great Powers cannot maintain access to oil supplies in order to keep their power projection capabilities at a similar level as other Great Powers, they lose their status of being one. With the historical examples of the UK, France and the Japanese case study, I showed how the maintenance of access to oil through balancing strategies is an inevitable part of being a tier 1 Great Power. The loss of their rank supported H3. The case of Russia showed that even with high accessibility to oil sources, it is an insufficient criterion to be considered a tier 1 Great Power
Leadership Competencies, Sustainable Strategies, and Energy Efficiency Initiatives for Driving Organizational Changes in the Healthcare Sector
Full text linkThis doctoral thesis investigates the interplay between leadership competencies, sustainable strategies, and energy efficiency initiatives in driving organizational change within the healthcare sector. The study provides a multidimensional perspective on how healthcare leaders navigate crises, implement sustainable strategies, and optimize energy consumption to enhance healthcare service delivery. Using a mixed-methods approach, the research offers empirical insights from Jordanian healthcare institutions, emphasizing the need for adaptive leadership in a rapidly evolving environment.
Healthcare organizations face numerous challenges, such as resource constraints, demographic shifts, technological advancements, and crises like COVID-19. Effective leadership is crucial for managing these changes, yet limited research has explored its impact on strategic decision-making and sustainability. This dissertation addresses this gap by analyzing leadership competencies, sustainable energy management, and crisis response strategies.
To provide a deeper understanding of the research, this dissertation presents three distinct papers, each exploring different aspects of leadership within the healthcare sector. Though varied in focus and methodology, these studies collectively contribute to the overarching theme of leadership dynamicity and its impact on healthcare transformation.
First Paper: Analysis of Leadership Competencies Based on Organizational Change: Case of Educational Hospitals in Jordan
• This quantitative study explores leadership characteristics in Jordanian educational hospitals, focusing on how demographic factors influence leadership competencies.
• Gender and age have a significant impact on leadership behavior, with leaders over 40 years old demonstrating higher levels of leadership competencies.
• Findings indicate that leaders exhibit key competencies such as integration, empathy, innovativeness, and adaptability.
• This study links leadership characteristics with effective change management, particularly in the context of organizational reform, offering valuable insights for improving leadership practices in healthcare settings.
Second Paper: Energy Efficiency in Healthcare Institutions
• This narrative review assesses energy efficiency strategies in healthcare settings, emphasizing the need for sustainable resource utilization.
• Best practices in energy management are identified, including the use of control algorithms, hybrid automatic voltage control (HAVC) systems, and renewable energy solutions (e.g., green hospitals).
• The paper highlights the importance of integrating leadership-driven sustainability measures to enhance energy conservation efforts.
Third Paper: Challenges and Opportunities in Healthcare Reforms in Pre-and Post-COVID-19 Crisis: A Case of Jordan
• A qualitative study investigates the impact of COVID-19 on Jordanian healthcare institutions and leadership responses.
• The study identifies major challenges such as staff shortages, increased workload, and financial constraints.
• It also highlights opportunities for reform, including enhanced infection control practices, improved staff training, and patient management innovations.
Key Contributions:
Drawing from the insights provided in the three papers, this dissertation offers several important contributions:
• Identified four key leadership characteristics crucial for managing organizational change in healthcare: integration, empathy, innovation, and adaptability.
• Demonstrated that demographic factors (age, gender) significantly influence leadership competencies and decision-making processes.
• Established a framework linking leadership practices to sustainable energy management in healthcare institutions.
• Provided empirical evidence supporting the use of renewable energy sources and advanced energy-saving technologies in healthcare settings.
• Highlighted how crises like COVID-19 can serve as catalysts for sustainable healthcare reforms, prompting improved management practices and resilience strategies.
• Proposed actionable recommendations for integrating leadership training into healthcare management to enhance adaptability and sustainability.
• Introduced a novel theoretical framework combining sustainability theory, resource-based view, and crisis management principles to explain leadership’s role in healthcare transformation.
• Provided policy recommendations for healthcare administrators and government bodies to improve energy efficiency and leadership training programs.
• Suggested a scalable leadership development model that can be adapted by healthcare institutions beyond Jordan, making the findings applicable to a broader international context
The Role of Attachment in the Switching Dynamics of Liquid Consumption
Full text linkThis dissertation examines how consumers’ feelings of attachment and psychological ownership impacts their switching behavior in the fast-paced digitalized world. Digitalization is an organic part of consumers’ lives and infuses their daily routine. Many of us start the day by scrolling through digital content on Instagram, checking the weather forecast right before getting out of bed. One uses a digital app to unlock a shared bike going to work. Others work simultaneously with the colleagues on a cloud server, order meals on the Foodora platform for the lunch break, book some tickets online for our weekend program with friends, and watch Netflix before sleeping after a busy day. By providing access almost to any good and service regardless of space or time, digitalization has left its imprint on consumers: consumers seem conditioned to get whatever, whenever and wherever their actual preference demands. Preferably immediately.
Digitalization has also changed the consumption landscape. It has transformed supply by easing market entry practically for everybody (see prosumers’ appearance) and resulted in the proliferation of competition. Alternatives come and go in one click, as consumers have a low switching cost and an infinite number of choices to satisfy their needs. Decisions have never before been so easy to change and lasted so short-term. These circumstances led to liquid consumption, an ephemeral, dematerialized and access-based form of consumption (Bardhi & Eckhardt, 2017), where consumers’ relationships to possession have liquefied, attachment has weakened and temporary access over possession gained ground.
Literature suggests that the relationship between consumers and alternatives has become loose and lacks the feeling of attachment, or even ownership (Bardhi & Eckhardt, 2017). But are consumers really not attached to their favorite video-on-demand service, where the library is tailored to our preferences? Do they really not feel their profile as their own possession? Is their dampening indeed linked to the emergence of liquid consumption and the acceleration of switching? This dissertation suggests that – consumers still feel attachment and ownership over digital services,
– thus, attachment does not hinder consumers from switching,
– while psychological ownership significantly impacts their switching behavior
Chapters from modeling mortality in Hungary [védés előtt]
Full text linkThis thesis attempts to stochastically model and forecast mortality in Hungary, and aims to support the development of the population projection model of the Hungarian Central Statistical Office. From the Lee–Carter model family, we have selected both those used to forecast mortality in a single population, and those used to study the mortality of different populations together. We focused on four topics, each of which can improve the modeling of mortality in Hungary. These four topics are: 1) regional mortality projection, 2) prediction of mortality by cause of death, 3) investigation of the time-dependence of age-varying parameters, i.e., rotation and their incorporation into the prediction of some mortality models, and 4) application of tree-based machine learning methods to mortality projections. For these research topics, we fitted different but related mortality models, always comparing their results and predictive accuracy, and selecting the best performing model. We used the Lee–Carter model as the main benchmark for all our research topics. The topics are also linked in that we have always fitted multi-population models. We used long-term time series to forecast Hungarian mortality rates by region, main cause of death and sex up to the year 2050
Understanding User and Investor Perspectives on Robo-Advisors Adoption in Fintech
Full text linkThe integration of Artificial Intelligence (AI) into financial services, particularly through AI-powered chatbots and Robo-advisors, is redefining customer interaction and service delivery in the FinTech sector. Despite their technological sophistication and growing global presence, user and investor adoption of Robo-advisors remains uneven, primarily due to persisting concerns around trust, data security, and privacy. This doctoral dissertation investigates the determinants influencing the acceptance and adoption of Robo-advisors, drawing on both user-centric and technology-driven perspectives. The research aims to fill a gap in existing literature by offering a comprehensive, theory-driven framework that explains how Robo-advisors can be successfully adopted in different sociotechnical environments—specifically within Pakistan and Hungary.
The study employs a mixed-methods approach. The qualitative phase consists of semi-structured interviews with 34 FinTech professionals in Pakistan and 15 industry experts in Hungary. These interviews, analyzed using Grounded Theory and Social Representation Theory (SRT), identify key thematic concerns and solutions regarding trust, security, privacy, user experience, and institutional credibility. NVivo-based coding reveals six primary challenge domains: data-security anxieties, consent and privacy management, institutional trust gaps, user-experience design, technology integration, and regulatory oversight. Findings underscore that transparency in data handling, visible regulatory compliance, and intuitive user interfaces are crucial to building trust in AI-powered financial tools.
To quantitatively validate these qualitative insights, the research proceeds with a Systematic Literature Review (SLR) and hypothesis-driven Structural Equation Modeling (SEM). The SLR analyzes 22 empirical studies, identifying relevant theoretical models such as the Technology Acceptance Model (TAM), DeLone & McLean Information Systems Success Model (ISSM), and the Unified Theory of Acceptance and Use of Technology 2 (UTAUT2). Constructs like perceived ease of use, perceived usefulness, system and information quality, perceived risk, customer experience, and trust were incorporated into four competing SEM models tested using SmartPLS and Adanco.
The SEM analysis confirms that perceived ease of use, usefulness, and trust positively influence users’ attitudes toward Robo-advisors, which in turn affects their intention to adopt. Moreover, perceived information, system, and service quality impact customer experience, which mediates the relationship with adoption intention. Moderating effects of perceived risk and demographic variables (age, gender, digital literacy) are also found to be significant. The results reveal that while utilitarian value is foundational, emotional factors—especially trust and perceived security—are critical for mass acceptance, particularly in emerging markets.
From a managerial standpoint, the findings offer actionable strategies for FinTech providers. In Pakistan, emphasis should be placed on education, transparency, and secure infrastructure to mitigate user fears about data misuse. In Hungary, where regulatory compliance and user interface sophistication are key, firms must invest in localized language interfaces, public regulatory disclosures, and responsive customer onboarding. The study also highlights the growing ethical implications of deploying Large Language Models (LLMs) in Robo-advisors, especially around algorithmic bias and privacy consent.
This dissertation makes multiple theoretical and practical contributions. It advances the academic understanding of FinTech adoption by proposing an integrative adoption framework grounded in both user behavior theories and information systems models. Practically, it provides FinTech developers, regulators, and service providers with a roadmap to increase customer satisfaction, mitigate ethical risks, and improve adoption rates. Future research could explore longitudinal changes in user behavior, the role of human-AI collaboration in financial advice, and the cross-cultural transferability of Robo-advisor systems
Essays in Behavioral Economics
Full text linkIn recent decades, behavioral and experimental economics have sig-nificantly influenced the broader discipline of economics. These sub-fields have revitalized interest in incorporating psychological dimen-sions into models of economic behavior. Beyond this conceptual shift, behavioral and experimental economics uses alternate meth-ods —namely: incentivized laboratory and field experiments—that allow for rigorous testing of theories and policy interventions be-fore rolling them out on large scale. Such experiments empower researchers to generate controlled data, moving beyond passive ob-servation of real-world phenomena and allowing for more rigorous causal analysis.
This thesis aims to use these tools in four topics. While the topics are loosely connected, it demonstrates that the instruments available for behavioral and experimental economics can be relevant in various research areas. Chapter 1 demonstrates the use of field experiments in a classroom environment. Specifically, together with Barna Bakó and Éva Holb, we explore the effect loss aversion has on student’s academic performance in a university setting. The key idea being that individuals perceive losses stronger than gains of equal size Kahneman and Tversky, 1979, reframing students’ evalu-ation system so that it emphasizes losses from not performing well or not completing tasks might lead to improved academic perfor-mance. While loss aversion was analysed in several academic set-tings (Apostolova-Mihaylova et al., 2015; Faulk et al., 2019; Levitt et al., 2016; McEvoy, 2016; Smith et al., 2019) , the results are het-erogeneous: some find no effect in general but heterogeneous gender effects (Apostolova-Mihaylova et al., 2015), some find no heteroge-neous effects but relatively large treatment effects (Faulk et al., 2019; McEvoy, 2016; Smith et al., 2019), and some find negative effects (Bies-Hernandez, 2012). While our first aim is to clarify some of these findings, the novelty of our research design is that it allows us to test the novelty-effect of loss-aversion framing. That is: it might be that students perform differently due to the unorthodox nature of grading; later during the semester, the might find out that this is indeed only a reframing of the evaluation system, and they do not truly lose points in the process. If this is true, we expect the treatment effect to decrease over time. To our knowledge, we are the first to answer this question
Insurgents, Rebel Governance and State-building in sub-Saharan Africa
Full text linkThe scholarship on rebel governance has largely focused on understanding how rebels govern during conflict. However, little is known about how rebel governance influences long-term state reconstruction after conflict. The main objective of this research is to address this gap.
The research also explores governance activities of rebel groups within two main periods: during conflict, and when groups transitioned to rulers, with a focus on the provision of public goods. Without delving too deeply into a discussion on goods, as this is not the focus of the dissertation, and without discussing every type of good, although they should be observed holistically, the research instead looks at the provision of health services. The research investigates rebel group cooperation with vaccination campaigns focusing on the selected cases during the periods indicated above as an indicator of “interest in governance”.1 The analysis is based on documented empirical evidence
From pen and paper to pixels. – Usability and User Acceptance of Digital Signatures: A Customer-Centric Study in Hungarian SMEs [védés előtt]
Full text linkHungary ranks 22nd in the EU Digital Economy and Society Index, with only 53.2% of SMEs demonstrating basic digital intensity, compared to 57.7% across the EU (Digital Decade Country Report 2024: Hungary, 2024).
This dissertation investigates the adoption of a foundation tool, digital signature, through a customer-centric lens, employing an exploratory mixed-methods design: a phone-assisted survey of 200 SMEs and 14 semi-structured expert interviews.
Logistic regression reveals that company size and sectoral orientation -not partner proximity- predict adoption (Nagelkerk