Διπλωματική εργασία--Πανεπιστήμιο Μακεδονίας, Θεσσαλονίκη, 2025.
(EL)
The purpose of the thesis is to provide a holistic research and examination into aviation ecosystem, and more detailed to the issue of flight delays and cancellations. The main focus is on the impact that machine learning models can have, by predicting successfully such events and extracting use- ful knowledge for improvement. The study begins with an introduction to the aviation industry, by citing also historical data and highlighting factors that may lead to flight disruptions. Deepens to the significance of a flight’s delay or cancellation, examining their influence on aspects like the impact to smooth airports’ operation, financial losses for the airlines and on top of that, inconve- niences for the passengers as well Moreover, exploring the policies and regulations that are in use regarding the flights and the rights of the travellers in case of any change in the schedule. The research proceeds in an in-depth descriptive analysis of the available dataset, uncovering useful insights but the same time raising critical questions, such as whether is possible and how to predict accurately and prevent these issues proactively rather than merely reacting to them. To address and give answers to the main scope of the thesis, after the data preparation and attribute generation process, multiple machine learning models was trained from the dataset, in order to assess their performance in forecasting events. The evaluation is considering among others, the scores of the accuracy, recall, execution time, that offer useful information about the effectiveness of each pre- diction model and determine their performance. Finally, the best-performing algorithm was iden- tified, which achieved great scores almost in every category that the models were assessed. These well-defined results, demonstrating its ability to achieve the highest efficient, while the same time convincing the users to trust the predicting models. In conclusion, the study tries to emphasize the importance, the capabilities and the impact that machine learning can have in real life problems. Data-driven solutions that can derive also from machine learning, are crucial in every business in nowadays. Equally crucial is the contribution of machine learning in tackling challenges and the potential applications of the findings to enhance operational efficiency in airlines and airports. Leveraging the insights derived from this research, all the stakeholders within the aviation indus- try can benefit, each applying the findings from their unique perspective and area of focus. The airlines can improve their services and their schedules. On the other hand, the airports can also use the study’s insights in order to investigate further the factors that cause the delays, in order to improve their services or manage better the gate or runway allocation to the airlines, while the passengers can have on-time information about their future flights. Overall, this study underscores the importance of predictive modelling in driving informed decision-making and improving the passenger experience in the aviation industry.
(EN)
Submitted by ΝΙΚΟΛΑΟΣ ΓΙΑΡΜΑΣ (
[email protected]) on 2025-02-14T13:45:33Z
No. of bitstreams: 2
license_rdf: 805 bytes, checksum: 4460e5956bc1d1639be9ae6146a50347 (MD5)
GiarmasNikosMsc2025.pdf: 3717070 bytes, checksum: 5b60f5c8b1db0ed3a57c985e85e20ae4 (MD5)
(EN)
Approved for entry into archive by Κυριακή Μπαλτά (
[email protected]) on 2025-02-17T14:15:30Z (GMT) No. of bitstreams: 2
license_rdf: 805 bytes, checksum: 4460e5956bc1d1639be9ae6146a50347 (MD5)
GiarmasNikosMsc2025.pdf: 3717070 bytes, checksum: 5b60f5c8b1db0ed3a57c985e85e20ae4 (MD5)
(EN)
Made available in DSpace on 2025-02-17T14:15:30Z (GMT). No. of bitstreams: 2
license_rdf: 805 bytes, checksum: 4460e5956bc1d1639be9ae6146a50347 (MD5)
GiarmasNikosMsc2025.pdf: 3717070 bytes, checksum: 5b60f5c8b1db0ed3a57c985e85e20ae4 (MD5)
Previous issue date: 2025
(EN)
University of Macedonia
