The aim of this notebook is the classification of observations of space in three different categories: stars, galaxies, quasars. To do so I use data from the Sloan Digital Sky Survey Classification.
The Sloan Digital Sky Survey is a project that periodically provides data from space observations free to the public. This project specifically uses data from the RD14 data release.
For this purpose a special 2.5 m diameter telescope was built at the Apache Point Observatory in New Mexico, USA. The telescope uses a camera of 30 CCD-Chips with 2048x2048 image points each. The chips are ordered in 5 rows with 6 chips in each row. Each row observes the space through different optical filters (u, g, r, i, z) at wavelengths of approximately 354, 476, 628, 769, 925 nm.
The dataset contains the following features:
- objid = Object Identifier
- ra = J2000 Right Ascension (r-band)
- dec = J2000 Declination (r-band)
- u = better of DeV/Exp magnitude fit
- g = better of DeV/Exp magnitude fit
- r = better of DeV/Exp magnitude fit
- i = better of DeV/Exp magnitude fit
- z = better of DeV/Exp magnitude fit
- run = Run Number --> maybe
- rereun = Rerun Number --> REMOVE
- camcol = Camera column --> maybe
- field = Field number --> maybe
- specobjid = Object Identifier
- class = object class (galaxy, star or quasar object)
- redshift = Final Redshift
- plate = plate number
- mjd = MJD of observation
- fiberid = fiber ID
Some features are remove as they consist of unique identifiers or are irrelevant to the identification of celestial objects:
objid, specobjid,fiberid,mjd,plate, rerun, run
The following color channels are also removed, as they are highly correlated with each other and another channel:
r, i, z
The algorithm chosen for the classification task is XGBoost Classifier.
Optimization is acheived with a grid search and the parameters of the best performing model are:
{'gamma': 1, 'learning_rate': 0.05, 'max_depth': 6, 'min_child_weight': 1, 'n_estimators': 100}
After optimization the scores for the train and test set are, respectively 0.998 and 0.99.