Data Science Workflow in Radar Data Analysis
Copyright (c) 2022 Madácsi Richárd
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.
Abstract
Data science is one of the hottest topics in the 21st century. The reason for this is probably the emergence of advanced machine learning algorithms based on neural networks, with which the possibilities seem to be endless. Therefore, those companies which do not want to be left behind, have to invest in this field heavily. However, most of the time the tasks that need to be done before applying machine learning algorithms do not get enough attention. These are data cleaning, filtering, transforming, feature engineering, which can affect the accuracy of the model more than the selection of the algorithm or its parameters. Quite a few tools are available for free, which makes the data science workflow efficient, although in data analysis focusing on ATM developing bespoke software is often necessary. The article aims to present the most common
requirements of that through examples and a small case study.
Keywords:
How to Cite
References
1. The Official Blog of Kaggle.com, Q&A with Xavier Conort. Online: http://blog.kaggle.com/2013/04/10/qa-with-xavier-conort/
2. C. Byrne, Development Workflows for Data Scientists. Sebastopol, California, O’Reilly Media, 2017.
3. Szarvas D., Tichy R., Rohács D., „Mesterséges intelligencia alkalmazása az aviatikában,” Repüléstudományi Közlemények, 31. évf. 1. sz. pp. 183–204. 2019. Online: https://doi.org/10.32560/rk.2019.1.15
4. P. Domingos, „A Few Useful Things to Know about Machine Learning,” Communications of the ACM, Vol. 55 No. 10. pp. 78–87. 2012. Online: https://doi.org/10.1145/2347736.2347755
5. Z. Wang, M. Liang, D. Delahaye, „Short-Term 4D Trajectory Prediction Using Machine Learning Methods,” In SID 2017, 7th SESAR Innovation Days, 2017. pp. 1–9. Online: https://www.sesarju.eu/sites/default/files/documents/sid/2017/SIDs_2017_paper_11.pdf
6. D. Cielen, A. D. B. Meysman, M. Ali, The Data Science Process, In Introducing Data Science. New York, Manning Publications, 2016.
7. F. Herrema, et al., „A Novel Machine Learning Model to Predict Abnormal Runway Occupancy Times and Observe Related Precursors,” In 12th USA/Europe Air Traffic Management Research and Development Seminar (ATM2017) 2017. pp. 1–11. Online: https://pure.tudelft.nl/ws/portalfiles/portal/31444878/12th_ATM_RD_Seminar_paper_107.pdf
8. Z. Wang, M. Liang, D. Delahaye, Automated Data-Driven Prediction on Aircraft Estimated Time of Arrival, SID 2018, 8th SESAR Innovation Days, 2018. pp. 1–8.
9. V. Kumar, L. Sherry, R. Kicinger, „Runway Occupancy Time Extraction and Analysis Using Surface Track Data,” In Transportation Research Board Annual Meeting, Transportation Research Board Paper, 10-3676, Washington, D.C., Jan. 2010.
10. S. Ayhan, P. Costas, H. Samet, „Predicting Estimated Time of Arrival for Commercial Flights,” In KDD ’18 Proceedings of the 24th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining London, 2018. pp. 33–42. Online: https://doi.org/10.1145/3219819.3219874
11. R. Madacsi, R. Markovits-Somogyi, „Bank Angle Estimation Using Radar Data,” Periodica Polytechnica Transportation Engineering, Vol. 47, No. 1. pp. 1–5, 2019. Online: https://doi.org/10.3311/PPtr.11653
12. C. O’Neil, R. Schutt, Doing Data Science: Straight Talk from the Frontline. Sebastopol, California, O’Reilly Media, 2013.
13. Madácsi R., Baráth M., Sándor Zs., A speciális térgeometriára támaszkodó „PointMerge” légiforgalmi irányítási módszer továbbfejlesztése. Budapest, IFFK, 2015.