Iron-Classifier is a project that focuses on classifying the Iron spectrum from the light nuclei in a supervised setting using various neural network model approaches. The project utilizes data sourced from the private Monte Carlo data for the MAGIC collaboration, specifically for heavy nuclei in cosmic rays.
-
ad: The main namespace of the project.
-
weights: contains the pre-trained weights of the various models.
The ad
namespace serves as the core module where the different models for classification are defined. The models
subpackage contains custom implementations of neural network models optimized for the specific task of classifying Iron spectra.
Additionally, the utils.py
module houses general plots and utility functions essential for data preprocessing, evaluation, and visualization.
To further enhance the project, the plots.py
file has been introduced to the structure, enabling the generation of additional plots and visualizations related to the classification process.
Feel free to modify, adjust, and expand the readme according to your specific needs and the details of your project.