Exoplanet Exploration

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. Three different machine learning models were created to help process this data; they are capable of classifying whether a record is a candidate exoplanet from the dataset of observed objects.

Programs Used:

• Python Pandas
• Scikit-Learn
• Keras

Resources

• Exoplanet Data Source

Steps to Creating the Models:

1. Preprocessed the raw data

   • Cleaned the data.
   • Removed unnecessary columns.
   • Scaled the data with the MinMaxScaler.
   • Separated the data into training and testing data.

2. Tuned the models

   • Used GridSearch to tune model parameters.
   • NOTE: Only used GridSearch on the SVM and Logistic Regression models.

3. Quantified the models

   • Found the best score from the trained GridSearch model.
   • Made predicitons from the hypertuned model.

Findings

I predicted that the Deep Learning model would have the highest accuracy because it used more than one layer of hidden nodes. The SVM and Logistic Regression models had around the same accuracy (0.87) once they were hypertuned with the GridSearch parameters. The Deep Learning model had a slightly higher accuracy, (0.89). The scores of the hypertuned GridSearch models slightly increased when compared to the test scores of the regular SVM and Logistic Regression models. Overall, the Deep Learning model seemed to be the most accurate.