Over a period of nine years in deep space, the NASA Kepler space telescope has been out on a planet-hunting mission to discover hidden planets outside of our solar system. This measurement data has been collected along with what the classification of the observation is.
We used the Exoplanet Data Source from Kaggle. Description of exoplanet dataset columns can be found here.
The following machine learning models, capable of classifying candidate exoplanets, were created from the above-mentioned dataset:
- Random Forest
- Decision Tree
- Gradient Boosting
- Logistic Regression
- Deep Learning
EDA, hyperparameter tuning and feature importance were performed for the first 4 models and for Deep Learning 2-layer and 3-layer models were used.
All models are saved and there is a separate notebook for each machine learning model.
- Tools/techniques used: Python, Jupyter Notebook, Scikit-learn, Pandas, Keras-Tensorflow, Matplotlib, Seaborn
GitHub Repository for this project.