This repository investigates SDSS DR-17 MaStar dataset of star spectra for the purposes of exploring statistical learning topics: AutoML, regression, and PCA. Also, in case it's not clear yet, I'm a bit of an astronomy geek ❤️😊
We will start by looking at pretty photos of some stars. Then we will check how regularized regression manages to capture details in the spectra, and how PCA-preprocessed regression compares to it. We will also look for an answer to the question: can a good prediction be made just based on the star colour we can see?
To make everything easier to comprehend, the notebooks are generously sprinkled with graphs and diagrams.
Start reading with the exploration notebook, and find out what mysterious and special star we have stumbled upon when browsing through the bad regression results ⭐!