COMPARISON AND ANALYSIS OF DIFFERENT MACHINE LEARNING METHODS ON ASTEROID DIAMETER PREDICTIONS BASED ON THE NASA SMALL CELESTIAL BODIES DATABASE

The database of small celestial bodies NASA is provided by the Jet Propulsion Laboratory and represents the collected information about asteroids and comets, describing their parameters available for observation and determination, including physical ones, as well as their classification and data on the number and duration of observation. Many of these celestial techs have an incomplete description of their properties, which makes it difficult to predict their behavior and potential interaction with other objects in space, including man-made ones. This study proposes a solution to a certain part of the problems of asteroid exploration by finding a prediction of the diameter of asteroids based on information from the NASA database and the results of machine learning methods on processed data from the source. For this research, some of the most commonly used algorithms for implementing such prediction models have been selected, such as KNN, linear regression, random forest, decision trees, and gradient boosting. Applied machine learning algorithms were evaluated based on the results of diameter prediction accuracy, speed of training and prediction process, and square mean error rates. The study will help to choose the most optimal approach for predicting this feature of asteroids, describe the process of data pre-processing, while achieving the best performance of the model, and analyze the correlations between the properties of these celestial bodies.

1. Alexandropoulos S.A., Kotsiantis S. and Vrahatis M. (2019) The Knowledge Engineering Review, 34, pp.1–33. https://doi.org/10.1017/S026988891800036X.

2. Altman E. (1968) The Journal of Finance, pp. 589–609.

3. Badescu. Asteroids: Prospective Energy and Material Resources. Springer Berlin, Heidelberg, 689 p.

4. Carruba V., Aljbaae S., Domingos R.C., Huaman M. and Barletta W. (2022) Celestial Mechanics and Dynamical Astronomy, 134, p. 36. https://doi.org/10.1007/s10569-022-10088-2.

5. Carruba V., Aljbaae S., Domingos R.C., Lucchini A. and Furlaneto P. (2020) Monthly Notices of the Royal Astronomical Society, 496(1), pp. 540–54. https://doi.org/10.1093/mnras/staa1463.

6. Chao H., Yue-hua M., Hai-bin Z. and Xiao-ping L. (2017) Chinese Astronomy and Astrophysics, 41(4), pp. 549– 557. https://doi.org/10.1016/j.chinastron.2017.11.006.

7. Chapman C. and Morrison D. (1994) Nature, 367, pp. 33–40. https://doi.org/10.1038/367033a0.

8. Dodge, The Concise Encyclopedia of Statistics, Springer, New York, 2008, 616 p.

9. Donnison J.R. and Sugden R.A. (1984) Monthly Notices of the Royal Astronomical Society, 210(3), pp. 673–682. https://doi.org/10.1093/mnras/210.3.673.

10. Emmanuel T., Maupong T. and Mpoeleng. (2021) Journal of Big Data, 8, 140 p. https://doi.org/10.1186/s40537021-00516-9.

11. Jet Propulsion Laboratory of California Institute of Technology, Small-Body Database Query. Retrieved May 3, 2023, from https://ssd.jpl.nasa.gov/tools/sbdb_query.html.

12. Kirch. Encyclopedia of Public Health, Springer, Dordrecht, 2008, 1600 p.

13. Kobayashi N., Oyamada Y., Mochizuki Y. and Ishikawa H., 14th IAPR International Conference on Machine Vision Applications (MVA) (Tokyo, 18-22 May 2015), p. 551–554.

14. Kotz S. and Johnson N. L. (1992) Breakthroughs in Statistics: Methodology and Distribution, Springer New York, NY, 600 p.

15. Lovric, International Encyclopedia of Statistical Science (Springer Berlin, Heidelberg), 79 p.

16. Sanchez P., Colombo C., Vasile M. and G. Radice. (2009) Journal of Guidance, Control and Dynamics, 32, pp. 121–142. https://doi.org/10.2514/1.36774.

17. Shang H., Wu X., Qiao D. and Huang X. (2018) Aerospace Science and Technology, 79, pp. 570–579. https://doi.org/10.1016/j.ast.2018.06.002.

18. Smirnov E.A. and Markov A.B. (2017) Monthly Notices of the Royal Astronomical Society, 469(2), pp. 2024–2031.https://doi.org/10.1093/mnras/stx999.

19. Steinbach M., Kumar V. and Tan P.-N. (2006) Introduction to Data Mining, Addison Wesley, Pearson, 165 p.

20. Wang, Y. (2023). Highlights in Science, Engineering and Technology, 39, pp. 201–208. https://doi.org/10.54097/hset.v39i.6527.