| |
 |
| Paper: |
AI in Astronomy |
| Volume: |
541, ADASS XXXIII |
| Page: |
165 |
| Authors: |
Maggie Lieu |
| DOI: |
10.26624/ZLAT9610 |
| Abstract: |
This study presents an automated classification system for identifying
galaxies, stars, and quasi-stellar objects (QSOs) using data from the Sloan Digital Sky
Survey (SDSS). Leveraging magnitudes across the u, g, r, i, and z filters along with redshift information, we employed a Random Forest classifier to discern between these
astronomical objects. The model achieved an impressive classification accuracy of
97.74% on a preprocessed dataset, underscoring the efficacy of machine learning techniques in astronomical classifications. Notably, our analysis revealed that while the
classifier effectively distinguished between stars and other objects, misclassifications
primarily occurred between galaxies and QSOs. This finding highlights the spectral
similarities between these two classes, particularly in active galactic nuclei. Furthermore, the study delves into the importance of individual features, with redshift and
specific magnitudes emerging as key discriminators. The results demonstrate the potential of applying advanced data analysis techniques to large astronomical datasets,
offering valuable insights into the nature and categorization of celestial objects. This
research contributes to the growing field of astrophysical data analysis, suggesting
pathways for further enhancement through feature engineering and advanced modeling
techniques. Disclaimer: This research (and paper) was written (largely) by ChatGPT
(http://openai.com). |
|
|
|
|
|
