Identifying galaxies, quasars, and stars with machine learning

This project was done as a part of Wolfram India Data Science School 2022. Under the supervision of Dr.Sotiris Michos. Here, I Attempted to classify galaxies, quasars, and stars (from the SDSS database) based on the photometric data of the telescopes with Decision Trees, Naive Bayes, Markov Chains, Gradient Boosted Trees, and Neural Network Algorithms. The results reflected the highest accuracy with Neural Networks with a specific netchain among all other algorithms.

Compared and analyzed the importance of the astronomical feature across all models with Shapley values and performed exploratory data analysis with the dataset to find patterns and anomalies.

Link to my project https://community.wolfram.com/groups/-/m/t/2443310