Doctoral Thesis

Μόνιμο URI για αυτήν τη συλλογήhttps://repository.ihu.gr/handle/11544/48406

Περιηγούμαι

A B C D E F G H I J K L M N O P Q R S T U V W X Y Z

Α Β Γ Δ Ε Ζ Η Θ Ι Κ Λ Μ Ν Ξ Ο Π Ρ Σ Τ Υ Φ Χ Ψ Ω

Τώρα δείχνει 1 - 3 από 3

Improved data science methods for real world business applications
Papageorgiou, Georgios
Data Science (DS) serves as a fundamental pillar across all industries, with businesses relying more than ever on data-driven insights for decision-making. Various methodologies are applied to industry data, often with multiple variations, aiming to extract valuable insights that can be translated by the corresponding stakeholders into actionable decisions. However, technological advancements and research innovations necessitate the exploration of new methodologies to drive progress in DS, Data Mining (DM), and Machine Learning (ML) research applied to industry data. This thesis sets out to explore three main pillars and introduce methods designed to enhance DS and DM applications within business contexts. To investigate and demonstrate these methods, we applied them to real-world business data from diverse domains, specifically basketball data, financial stock data, and energy-related data. Our primary research questions are focused on exploring, introducing, and presenting advanced methods for applying DS and DM techniques to real-world business data, including strategies for data retrieval. This thesis introduces innovative ML modeling approaches, new business-oriented Key Performance Indicators (KPIs) for method evaluation, novel ML evaluation methods rooted in DS fundamentals, advanced techniques for anomaly detection, and the application of Association Rule Mining (ARM) to clean datasets. Additionally, we present a new innovative normalization method for time-series data. This thesis first addressed advancements in individual modeling for highly volatile sports data, examining how optimal results can be achieved relative to standard ML methodologies. This effort aims to benchmark the forecasting performance of 14 ML models based on 18 advanced basketball statistics and KPIs. Additionally, business-oriented target variables in basketball are forecasted, focusing on minimizing forecasting errors for volatile data. A new business-oriented evaluation metric called Weighted Average Percentage Error (WAPE) is introduced. Additionally, an ensemble of anomaly detection methods was applied, including the Density-Based Spatial Clustering of Applications with Noise (DBSCAN) algorithm. The Apriori algorithm for ARM is also implemented to identify associations among different types of data. Lastly, we introduced a novel normalization method for time-series data called Adaptive Sliding Window Normalization (ASWN). This method dynamically adjusts normalization window sizes, based on anomalies detected through multiple methods. For finalizing these anomalies, DBSCAN is employed, and the Akaike Information Criterion (AIC), in conjunction with AutoRegressive Integrated Moving Average (ARIMA) models, are used to determine optimal window sizes in the absence of anomalies. By presenting all these advances in DS, spanning data processing, forecasting, and pattern and association recognition, this work aims to contribute to the fields of DS, DM, and ML as applied to real-world data and scenarios. All methods introduced are evaluated and discussed in terms of their applications, highlighting their contributions to the DS field, as well as their applicability in real-world business contexts. The primary focus remains on advancing techniques for extracting valuable insights from diverse types of raw data. Additionally, further research objectives are outlined to explore these methods and processes further, aiming to achieve optimal knowledge extraction and continued improvements in the field.
Improving the efficiency of the catalytic hydrogenation of CO2 via intensified process concepts and novel catalysts
Koidi, Vasiliki
The on-going use of fossil fuels leads to the persistent increase of carbon dioxide emissions, which in turn contribute to global warming and climate change. To mitigate this threat, a substantial part of research has been devoted to CO2 conversion into value-added products such as methanol or dimethyl ether. Methanol synthesis via CO2 hydrogenation reaction is exothermic, favored at high pressures and low temperatures. However, CO2 is a fully oxidized, thermodynamically stable, and chemically inert molecule that requires relatively high activation temperature. The reaction is therefore limited due to thermodynamics at high temperature and due to kinetics at low temperature. The purpose of the present thesis is to improve the efficiency of CO2 hydrogenation to value-added products, such as methanol or dimethyl ether, by developing novel catalysts and intensified process concepts with special emphasis on the experimental development of the sorption-enhanced (SE) methanol synthesis process. On the catalyst side, Mo₂C-based catalysts were systematically studied to address kinetic constraints, revealing that both the catalyst preparation method and Cu promotion significantly affect the catalytic performance, while Cu addition was also shown to increase methanol selectivity at temperatures below 275 °C. Additionally, the stability of a commercial CuO/ZnO/Al₂O₃ catalyst exposed to typical off-gases impurities was investigated, showing that activity loss is primarily due to a reduction in exposed Cu surface area caused by the poisons. To address thermodynamic limitations, the SE methanol synthesis concept was experimentally validated using zeolite-based sorbents for continuous water removal, enhancing methanol yield by up to 115%. The impact of different zeolites on performance and stability was evaluated, revealing that zeolite 4A offers the highest stability. Another route to circumvent thermodynamic limitations is to further convert the methanol produced to dimethyl ether. To that end, 3D-printing technology was employed to prepare structured acidic supports for methanol dehydration and bifunctional catalysts for the one-step CO₂-to-DME conversion, achieving high stability and selectivity.
The Innovation Engine Inside Government The Role of Labs in Public Sector Digital Transformation & Value Creation
(ΔΙΠΑΕ, 2026-06-08) Δεληγιάννης, Αθανάσιος; Περιστέρας, Βασίλειος; Σχολή Επιστήμης και Τεχνολογίας, Τμήμα Επιστήμης και Τεχνολογίας
The accelerating digital transformation of public administration demands a redefinition of how governments create public value. While private sector organizations have long relied on research and development laboratories to generate innovation, the public sector’s approach has historically been fragmented, incremental, and often constrained by bureaucratic norms. This thesis investigates Public Sector Innovation (PSI) units as institutional mechanisms that enable governments to transition from hierarchical, rule-based systems to open, adaptive, and citizen-centered models of governance. Set within the broader context of open government and digital democracy, the study argues that digital transformation is not merely a technical challenge but a profound organizational and cultural shift. It requires public administrations to embrace participatory methods, transparency, and cross-sector collaboration while maintaining accountability and privacy. The work explores how innovation labs function as vehicles of change, promoting experimentation, co-creation, and collaborative problem-solving within government structures. Thus, the central objective of the thesis is to examine the role of PSI units in generating innovation and public value, identifying the characteristics, success factors, and challenges that define their effectiveness. To achieve this, the research adopts a multi-layered methodology: An extensive literature review compares innovation paradigms across public and private sectors, highlighting typologies, drivers, and barriers of innovation. Comparative case studies of nine PSI units worldwide—ranging from MIT’s Connection Science Living Lab to the EU’s Joint Research Centre Policy Lab and the OECD’s OPSI—map their organizational models, missions, and practices. In-depth qualitative analysis of two atypical cases, the Greek eGovernment Team and the Norwegian Brønnøysund Register Center, provide insights into innovation processes within distinct national contexts. Data from these cases were analyzed thematically supported by inductive coding and iterative validation between researchers. The findings are integrated into a framework for assessing the institutional maturity and sustainability of PSI units. The study reveals that successful PSI units combine strategic alignment with political priorities and organizational autonomy to enable experimentation. Their effectiveness rests on five main pillars: Mission clarity and adaptability: Innovation units with well-defined yet flexible missions—co-created with stakeholders—exhibit higher resilience and policy relevance. Learning-oriented culture: Acknowledging and learning from failure strengthens organizational agility and knowledge sharing. Human-centered capability building: Staff diversity and continuous professional development are vital to sustain innovation momentum. Collaborative ecosystems: Effective innovation units function as intermediaries connecting government, academia, civil society, and business. Open communication and visibility: Transparency and active dissemination of knowledge increase legitimacy and stakeholder engagement. Common success factors include policy integration, prototype adoption, and capacity-building outcomes - not so commercial results. The research identifies fragility and political dependency as recurring challenges—many labs face closure following leadership or funding changes. Institutional embedding, long-term mandates, and evaluation frameworks are crucial to overcoming these vulnerabilities This research concludes that PSI units should not be considered isolated innovation outposts but rather systemic enablers of digital governance. Their value lies in institutional learning, network building, and the democratization of innovation processes. Public sector innovation is increasingly characterized by collaborative, iterative, and citizen-driven practices, transforming governments into facilitators of collective intelligence and public value creation. The thesis also provides an evaluative framework and typology for PSI units, offering policymakers and managers practical guidance for sustaining innovation cultures within bureaucratic systems. It also outlines practical scenarios that can guide the creation and success of such units under different conditions

Περιηγούμαι

Πλοήγηση Doctoral Thesis ανά Τίτλο