Çarpışma Riski Bulunan Asteroitlerin Makine Öğrenmesi ile Tespiti

Öz

Asteroitler geçmişten günümüze kadar insanların dikkatini çekmektedir. Kadim medeniyetlerin inanç ve kültürlerinde de geniş yer almaktadır. İnsanoğlunun keşfetme ve merak duygusu bu cisimlere olan ilgisinin artmasına neden olmaktadır. Teknolojinin belirli bir seviyeye gelmesiyle asteroitlerin tespiti, teşhisi ve materyalleri net bir şekilde bulunabilmektedir. Bu cisimlerin izleyecekleri rota ve çarpışma etkileri sürekli olarak gözlem gerektirmektedir. Çalışmamızda Kaggle’da bulunan ve kaynağı NASA-JPL olan bir asteroit veri seti kullanılarak Dünya’ya çarpma ihtimali olan asteroitlerin sınıflandırılması yapılmıştır. Veri setinde 4687 asteroit verisi bulunmaktadır. Veriler üzerinde eksik verileri doldurulması, anomali tespit etme ve normalizasyon gibi ön işleme aşamaları uygulanmıştır. Daha sonra korelasyon yardımıyla tehlikelilik durumları için verisetinden 19 adet öznitelik tespit edilmiştir. Öznitelikler ile Karar Ağacı, Naive Bayes, Lojistik Regresyon, Rastgele Orman, Destek Vektör Makineleri, K-En Yakın Komşu, Xgboost ve Adaboost makine öğrenmesi algoritmaları kullanılarak asteroit sınıflandırması yapılmıştır. Farklı nöron ve katman sayılarına sahip yapay sinir ağı ile veriler eğitilmiş ve sınıflandırma algoritmaları ile karşılaştırılmıştır. Karşılaştırma sonucunda en yüksek başarımı %99.80 ile AdaBoost algoritması ile sağlanmıştır. Çalıştırılan tüm sınıflandırma algoritmalarında ızgara-arama yöntemi kullanılarak hiperparametre optimizasyonu yapılmıştır. Böylelikle sürekli gözlem gerektiren ve yüksek miktardaki verilerin daha performanslı bir şekilde işlenmesini sağlayan bir yöntem önerilmiştir.

Anahtar Kelimeler

• Asteroit Sınıflandırması
• Anomali Tespiti
• Kara Ağaçları
• Yapay Sinir Ağları

Machine Learning Detection of Collision-Risk Asteroids

Öz

Asteroids have attracted people's attention from the past to the present. It has a wide place in the beliefs and cultures of ancient civilizations. The sense of discovery and curiosity of human beings causes an increase in their interest in these objects. With the technology coming to a certain level, the detection, diagnosis and materials of asteroids can be found clearly. The route and collision effects of these objects require constant observation. In our study, asteroids that are likely to hit the Earth have been classified using an asteroid data set in Kaggle and the source of which is NASA-JPL. The dataset contains 4687 asteroid data. Pre-processing steps such as filling in missing data, anomaly detection and normalization were applied on the data. Then, with the help of correlation, 19 features were determined from the dataset for dangerous situations. Asteroid classification was made by using Decision Tree with features, Naive Bayes, Logistic Regression, Random Forest, Support Vector Machines, K-Nearest Neighbor, Xgboost and Adaboost machine learning algorithms. With the artificial neural network with different number of neurons and layers, the data were trained and compared with classification algorithms. As a result of the comparison, the highest performance was achieved with the AdaBoost algorithm with 99.80%. Hyperparameter optimization was performed using the grid-search method in all the classification algorithms that were run. Thus, a method that requires continuous observation and enables the processing of large amounts of data in a more efficient way has been proposed.

Anahtar Kelimeler

• Asteroid Classification
• Anomaly Detection
• Decision Tree
• Artificial Neural Networks

Kaynakça

1. Furfaro, R., Barocco, R., Linares, R., Topputo, F., Reddy, V., Simo, J., Le Corre, L., "Modeling irregular small bodies gravity field via extreme learning machines and Bayesian optimization", Advances in Space Research, 2021, 67(1): 617-638.
2. Kabaş, A., Bulut, İ., Doğru, S. S., Akin, T., "Bazı Ana Kuşak Asteroidlerin Işık Eğrileri Ve Işık Eğrilerinden Belirlenen Parametreler”, XVI. Ulusal Astronomi Kongresi ve V. Ulusal Öğrenci Astronomi Kongresi, Çanakkale, 1188-1199, 2008.
3. Chapman, C. R., Ramlose, T., "Solar system exploration" NASA STI/Recon Technical Report N, 1989, 89: 25944.
4. Popescu, M., Licandro, J., Carvano, J. M., Stoicescu, R., de León, J., Morate, D., Boacă, I. L., Cristescu, C. P., "Taxonomic classification of asteroids based on MOVIS near-infrared colors", Astronomy & Astrophysics, 2018, 617(A12).
5. Cambioni, S., Bennett, C. A., Walsh, K. J., DellaGiustina, D. N., Golish, D. R., Becker, K. J., Lauretta, D., S., "A search for smooth terrains on asteroid (101955) Bennu using machine learning", In EPSC-DPS Joint Meeting 2019, 2019.
6. Erasmus, N., Mommert, M., Trilling, D. E., Sickafoose, A. A., Van Gend, C., Hora, J. L., "Characterization of near-earth asteroids using KMTNET-SAAO", The Astronomical Journal, 2017, 154(4): 162.
7. Heinze, A. N., Tonry, J. L., Denneau, L., Flewelling, H., Stalder, B., Rest, A., Smith, K. W., Smartt, S. J., Weiland, H., "A first catalog of variable stars measured by the Asteroid Terrestrial-impact Last Alert System (ATLAS)", The Astronomical Journal, 2018, 156(5): 241.
8. Smirnov, E. A., Markov, A. B., "Identification of asteroids trapped inside three-body mean motion resonances: a machine-learning approach", Monthly Notices of the Royal Astronomical Society, 2017, 469(2): 2024-2031.

9. Mommert, M., Trilling, D. E., Hora, J. L., Lejoly, C., Gustafsson, A., Knight, M., Moskovitz, N., Smith, H. A., "Systematic characterization of and search for activity in potentially active asteroids", The Planetary Science Journal, 2020, 1(1): 10.
10. Beasley, M., Lewicki, C. A., Smith, A., Lintott, C., Christensen, E., "AsteroidZoo: A new zooniverse project to detect asteroids and improve asteroid detection algorithms", In AGU Fall Meeting Abstracts, 2013.
11. Nugent, C. R., Dailey, J., Cutri, R. M., Masci, F. J., Mainzer, A. K., "Machine learning and next-generation asteroid surveys", In AAS/Division for Planetary Sciences Meeting Abstracts, 2017, 49: 103-03.
12. Rabeendran, A. C., Denneau, L. "A two-stage deep learning detection classifier for the atlas asteroid survey", Publications of the Astronomical Society of the Pacific, 2021, 133(1021): 034501.
13. Pasko, V., "Prediction of Orbital Parameters for Undiscovered Potentially Hazardous Asteroids Using Machine Learning", In Stardust Final Conference, Springer Netherlands, 52: 45-65.
14. Johnson, C. A., DellaGiustina, D. N. "Hazards on Hazards, Ensuring Spacecraft Safety While Sampling Asteroid Surface Materials", In AGU Fall Meeting Abstracts, 2016, 2016: NH13A-1748.
15. Gustetic, J. L., Friedensen, V., Kessler, J. L., Jackson, S., Parr, J., "NASA's Asteroid Grand Challenge: Strategy, Results, and Lessons Learned", Space Policy, 2018, 44: 1-13.
16. Lieu, M., Baines, D., Giordano, F., Merin, B., Arviset, C., Altieri, B., Conversi, B., Carry, B., "Deep Learning of Astronomical Features with Big Data", Astronomical Data Analysis Software and Systems XXVII, 2019, 523: 49.
17. Fluke, C. J., Jacobs, C., "Surveying the reach and maturity of machine learning and artificial intelligence in astronomy", Wiley Interdisciplinary Reviews: Data Mining and Knowledge Discovery, 2020, 10(2): e1349.
18. Dotto, E., Della Corte, V., Amoroso, M., Bertini, I., Brucato, J. R., Capannolo, A., ... & Fretz, K., "LICIACube-the Light Italian Cubesat for Imaging of Asteroids in support of the NASA DART mission towards asteroid (65803) Didymos", Planetary and Space Science, 2021, 199: 105185.
19. McIntyre, K. J., "Applying Machine Learning To Asteroid Classification Utilizing Spectroscopically Derived Spectrophotometry", Master Thesis, The University of North Dakota, 2019.
20. Tholen, D. J., Barucci, M. A., "Asteroid taxonomy", Asteroids II, 1989, 298-315.
21. Tedesco, E. F., Williams, J. G., Matson, D. L., Veeder, G. J., Gradie, J. C., Lebofsky, L. A., "Three-parameter asteroid taxonomy classifications", Asteroids II, 1989, 1151-1161.
22. Tholen, D. J., "Asteroid taxonomic classifications", In Asteroids II, 1989, 1139–1150.
23. Safavian, S. R., Landgrebe, D., "A survey of decision tree classifier methodology", IEEE transactions on systems, man, and cybernetics, 1991, 21(3): 660-674.
24. Rish, I., "An empirical study of the naive Bayes classifier", In IJCAI 2001 workshop on empirical methods in artificial intelligence, 2001, 3(22): 41-46.
25. Sperandei, S., "Understanding logistic regression analysis"; Biochemia medica, 2014 24(1): 12-18.
26. Breiman, L., "Random forests", Machine learning, 2001, 45(1): 5-32.
27. Cortes, C., Vapnik, V., "Support-vector networks", Machine learning, 1995, 20(3): 273-297.
28. Pal, M., "Random forest classifier for remote sensing classification", International journal of remote sensing, 2005, 26(1): 217-222.
29. Guo, G., Wang, H., Bell, D., Bi, Y., Greer, K., "KNN model-based approach in classification", In OTM Confederated International Conferences" On the Move to Meaningful Internet Systems", Berlin, 2003, 986-996.
30. Chen, T., Guestrin, C., "Xgboost: A scalable tree boosting system", In Proceedings of the 22nd acm sigkdd international conference on knowledge discovery and data mining, 2016, 785-794.
31. Chen, T., He, T., Benesty, M., Khotilovich, V., Tang, Y., Cho, H., Chen, K., "Xgboost: extreme gradient boosting", R package version 0.4-2, 2015, 1(4): 1-4.
32. Freund, Y., Schapire, R. E., "Experiments with a new boosting algorithm", icml, 1996, 96: 148-156.
33. Schapire, R. E., "Explaining adaboost", In Empirical inference, Berlin, 2013, 37-52.
34. Öztemel, E., "Yapay sinir ağlari", PapatyaYayincilik, İstanbul, (2003).
35. Şen, Z., "Yapay sinir ağları", Su Vakfı, İstanbul, (2004).