This file contains tools used for conducting machine learning tasks in Liquid Scintillator Detector. The purpose of 4 folders are listed below:
Preprocessing: Converting Monte Carlo Simulation .root file to a 2D Grid (theta, phi), and store them as a sparse matrix.
Network: Storing the Convolutional Neural Network Used to conduct the classification task, also include the hyperparameter search algorithm.
Analyzer: data analysis tool for making the pressure map and roc curves.
data: Data processing script to be use on BU SCC cluster.
Adding Spherical CNN feature modified from https://github.com/jonas-koehler/s2cnn