Harokopio University
An intelligent framework for the techno-economic analysis on Cloud services
Abstract
dc:descriptionThe fast growing environment of cloud services has produced a complicated price web that causes hurdles for consumers searching for cost-effective solutions. Within the context of this intricate landscape, this thesis presents a novel contribution known as the PriceOPS Cloud Costing Framework. The framework, which has been developed as a multi-methodological approach, operates on three axes and may be utilized across many product categories, including IaaS, CaaS, and PaaS. These categories are characterized by a comprehensive range of shared as well as unique attributes. The framework was designed based on three axes: the examination of pricing policy, the analysis of the comparative cost of cloud services, and the assessment of the growth of cloud service prices. In order to execute the task, a collection of heuristic methods has been integrated within a unified programming environment to construct a price index employing hedonic indices. Additionally, clustering techniques have been employed to compare prices of products offered by various providers, while machine learning algorithms have been utilized to analyze historical data and identify price trends. The selection of certain methodologies for addressing specific problems is guided by existing scholarly literature and established best practices in relevant fields, while also allowing for the potential utilization of other approaches and techniques that align with these lines of study. The primary focus of this study focuses around providing cloud service customers with a methodical framework for selecting IaaS, CaaS, and PaaS products, while considering the current pricing policy and its potential future developments. The key approaches incorporated in the framework have the potential to be modified and used for pricing strategies in many industries, thereby extending its application and influence. The study introduces intriguing possibilities for future research endeavors, focusing on enhancing algorithms to effectively operate in real-time data streaming scenarios. The framework has been intentionally built and developed with scalability in mind, allowing for its adaptability in addressing price difficulties not just in cloud computing but also in other domains, such as the Internet of Things (IoT).Within the framework of the study conducted for this thesis, the identified contribution is to the standardization of products across providers. The purpose is to enhance the efficiency and ease of comparing these products. In the field of price forecasting, investigating the packages of reserved instances of IaaS, the thesis introduces this approach. The accessibility and repeatability offered by the creation of open source libraries are considered to be significant advantages in the realm of research. By making both the code and datasets openly accessible, it facilitates the replication of the suggested methodologies and tools by academics from several fields, including the Internet of Things.
Degree
thesis:*- Grantor dc:publisher
- Harokopio University
- Year dc:date
- 2023
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Fragkiadakis, Georgios
- Φραγκιαδάκης, Γεώργιος
Subjects
dc:subject × 16- Υπολογιστική νέφους
- Μηχανική μάθηση
- Τεχνο‐οικονομική ανάλυση
- Κοστολόγηση
- Επιχειρισιακή απόφαση
- Cloud computing
- Machine learning
- Technoeconomical analysis
- Pricing optimization
- Consumer decision‐making
- Επιστήμη Ηλεκτρονικών Υπολογιστών και Πληροφορική
- Φυσικές Επιστήμες
- Επιστήμη ηλεκτρονικών υπολογιστών και Πληροφορική, άλλοι τομείς
- Computer and Information Sciences
- Natural Sciences
- Computer and Information sciences, miscellaneous
Rights
- Language dc:language
- eng
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/55177