Models for forecasting flight delays
The article deals with the problem of improving the operation of airports and aircrafts. It is proposed to use machine learning models and technologies to predict flight delays. Much of the effort was directed at collecting qualitative data on both air travel and the factors that could potentially a...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Igor Sikorsky Kyiv Polytechnic Institute
2023-12-01
|
| Series: | Adaptivni Sistemi Avtomatičnogo Upravlinnâ |
| Subjects: | |
| Online Access: | http://asac.kpi.ua/article/view/292243 |
| _version_ | 1797351365006065664 |
|---|---|
| author | D. Tarasonok Y. Oliinyk T. Likhouzova |
| author_facet | D. Tarasonok Y. Oliinyk T. Likhouzova |
| author_sort | D. Tarasonok |
| collection | DOAJ |
| description | The article deals with the problem of improving the operation of airports and aircrafts. It is proposed to use machine learning models and technologies to predict flight delays. Much of the effort was directed at collecting qualitative data on both air travel and the factors that could potentially affect it. Thanks to this, a dataset of almost half a million flights
was formed. The purpose of the work is to predict flight delays, which was done in both quantitative (delay for how many minutes) and qualitative (delay exceeds 15 minutes) options. 5 regression and 5 classification models of three different types were built to predict departure delays at the Atlanta airport, USA. To evaluate the effectiveness of the proposed models, several different quality measures were used, which reflect the expediency of using these models in terms of the needs
of each task. For the best model, the median absolute error is 5 minutes, which is an excellent result in predicting flight departure delays. Accurately predicting flight delays can provide guidance to logistics companies to more accurately plan their transportation, and in this industry, this is one of the main points in making a profit.
Ref. 20, pic. 13 |
| first_indexed | 2024-03-08T12:59:28Z |
| format | Article |
| id | doaj.art-83b9da4d1e454530aa956e99fc021013 |
| institution | Directory Open Access Journal |
| issn | 1560-8956 2522-9575 |
| language | English |
| last_indexed | 2024-03-08T12:59:28Z |
| publishDate | 2023-12-01 |
| publisher | Igor Sikorsky Kyiv Polytechnic Institute |
| record_format | Article |
| series | Adaptivni Sistemi Avtomatičnogo Upravlinnâ |
| spelling | doaj.art-83b9da4d1e454530aa956e99fc0210132024-01-19T12:06:17ZengIgor Sikorsky Kyiv Polytechnic InstituteAdaptivni Sistemi Avtomatičnogo Upravlinnâ1560-89562522-95752023-12-01243203310.20535/1560-8956.43.2023.292243330530Models for forecasting flight delaysD. Tarasonok0Y. Oliinyk1T. Likhouzova2Igor Sikorsky Kyiv Polytechnic InstituteIgor Sikorsky Kyiv Polytechnic InstituteIgor Sikorsky Kyiv Polytechnic InstituteThe article deals with the problem of improving the operation of airports and aircrafts. It is proposed to use machine learning models and technologies to predict flight delays. Much of the effort was directed at collecting qualitative data on both air travel and the factors that could potentially affect it. Thanks to this, a dataset of almost half a million flights was formed. The purpose of the work is to predict flight delays, which was done in both quantitative (delay for how many minutes) and qualitative (delay exceeds 15 minutes) options. 5 regression and 5 classification models of three different types were built to predict departure delays at the Atlanta airport, USA. To evaluate the effectiveness of the proposed models, several different quality measures were used, which reflect the expediency of using these models in terms of the needs of each task. For the best model, the median absolute error is 5 minutes, which is an excellent result in predicting flight departure delays. Accurately predicting flight delays can provide guidance to logistics companies to more accurately plan their transportation, and in this industry, this is one of the main points in making a profit. Ref. 20, pic. 13http://asac.kpi.ua/article/view/292243machine learningforecastingregression problemclassification problem |
| spellingShingle | D. Tarasonok Y. Oliinyk T. Likhouzova Models for forecasting flight delays Adaptivni Sistemi Avtomatičnogo Upravlinnâ machine learning forecasting regression problem classification problem |
| title | Models for forecasting flight delays |
| title_full | Models for forecasting flight delays |
| title_fullStr | Models for forecasting flight delays |
| title_full_unstemmed | Models for forecasting flight delays |
| title_short | Models for forecasting flight delays |
| title_sort | models for forecasting flight delays |
| topic | machine learning forecasting regression problem classification problem |
| url | http://asac.kpi.ua/article/view/292243 |
| work_keys_str_mv | AT dtarasonok modelsforforecastingflightdelays AT yoliinyk modelsforforecastingflightdelays AT tlikhouzova modelsforforecastingflightdelays |