Project Overview
This project focuses on the classification of exoplanets using data from NASA’s TESS mission. Our team developed a deep learning neural network capable of analyzing celestial objects and categorizing them into three classes: Confirmed Planets (CP), Planet Candidates (PC), and False Positives (FP).
My Role
I was responsible for designing, training, and validating the neural network models, optimizing them for high accuracy and real-time performance. I also implemented preprocessing pipelines and data augmentation strategies to enhance model performance on limited datasets.
Technologies Used
Python, TensorFlow/Keras, NumPy, Pandas, Jupyter Notebook, NASA TESS dataset, Data Augmentation, Neural Network Design, Model Evaluation Metrics.
Results & Achievements
The model achieved high classification accuracy and helped our team secure 2nd place in the NASA Space Apps 2025 challenge. The project highlights the application of deep learning techniques in astrophysics data analysis.
