Exoplanets_Finder

This project explores the dataset, and uses DL model to perform predictions

Dataset https://www.kaggle.com/datasets/keplersmachines/kepler-labelled-time-series-data

Dataset Summary: The Search for New Earths (Kepler Exoplanet Dataset)

Overview

This dataset originates from the NASA Kepler space telescope and records the change in flux (light intensity) of thousands of stars over time. The primary goal is to identify exoplanet-hosting stars by detecting periodic dimming, which suggests the presence of an orbiting planet. Each star is assigned a binary label:

2 (Exoplanet-Star): At least one confirmed exoplanet detected.
1 (Non-Exoplanet-Star): No confirmed exoplanet detected.

The dataset has undergone preprocessing, with data mainly derived from Kepler’s Campaign 3 observations, complemented by exoplanet data from other campaigns to increase the number of positive cases.

Dataset Structure

Train Set:

Rows (Observations): 5,087 stars
Columns (Features): 3,198 flux values (time-series data) + 1 label column
Class Distribution:

-37 Exoplanet-Stars
-5,050 Non-Exoplanet-Stars

Test Set:

Rows (Observations): 570 stars
Columns (Features): 3,198 flux values (time-series data) + 1 label column
Class Distribution:

-5 Exoplanet-Stars
-565 Non-Exoplanet-Stars