Exoplanet Identification Through Different Types of Deep-Learning Neural Networks

AJAS · 2020 Physics and Astronomy (inferred)

Sanskriti Deva · and Pallavi Maladkar, North Carolina Academy of Science

Overview

Due to the vast amount of data collected by the Kepler telescope, it takes a lot of time and work for scientists to go through each light curve and figure out whether or not the curve is a possible candidate to be an exoplanet. We decided to address this using neural networks. In this study, we created and tested two different types of neural networks, recursive and convolutional, to find which one was better in terms of speed and accuracy when identifying exoplanet light curves. We created, trained, and then tested the neural networks multiple times using the Kepler dataset. We were able to create a neural network that had an accuracy of about 93% (convolutional) and is quicker than previous ones. We found that on average the convolutional neural network was more accurate while the recursive was faster but not as accurate.

Competition history

AJAS 2020 Category not listed

Related projects

ISEF · 2025

Optimizing Exoplanet Detection: A Comparison of Neural Networks and Traditional Machine Learning Models on Light Curve Data

JSHS · 2023

Using Deep Neural Networks to Identify Exoplanets

ISEF · 2023

A Novel Method in Improving the Accuracy of Exoplanet Classification Using Machine Learning

ISEF · 2025

Innovative Machine Learning Approaches for Exoplanet Detection and Anomaly Identification: Applying Computational Techniques to Uncover Hidden Patterns in TESS Data

ISEF · 2018

Challenging Limitations: Using Deep Learning, Time Series Analysis, and Statistical Methods for Noise Reduction to Develop an Innovative Approach to Exoplanet Candidate Detection Using Earth-Based Telescopes

AJAS · 2026

Browse more like this

Source: AAAS Annual Meeting (Confex) / American Junior Academy of Science