This project focuses on detecting solar flares by analyzing solar images using a deep learning model. A solar flare is a sudden burst of radiation from the sun, releasing significant magnetic energy. The model classifies images into two categories: flare or non-flare.
The goal is to determine whether a solar flare has occurred based on an image of the sun.
The following preprocessing techniques were applied to the images:
- Resizing
- Normalization
- Data Augmentation
- Color Maps for Luminosity
You can download the dataset from the following link and change the path in the ipynb file:
https://drive.google.com/drive/folders/18AFYzmiGkwGw8GKTDxvnVnMNzQqMrZOI?usp=drive_link(#)
A Convolutional Neural Network (CNN) was used to train the model for flare detection.
The project also includes a detailed post-analysis of the model's performance, covering:
- Accuracy
- Precision
- Recall
- Confusion Matrix
- Area Under the Curve (AUC) These metrics were used to evaluate and fine-tune the model for better performance.