WSU efforts to vertex reconstruction in NOvA
Original Author: Michael Dolce mdolce@fnal.gov
Updated: Nov. 2023
This repository is the home of the NOvA Neutrino Interaction Vertexing project at Wichita State University.
The objective of the NOvA vertexing is to improve the NOvA reconstruction algorithms of particle interactions within the detector to improve energy reconstruction. This improved energy reconstruction, in turn, can help to improve a constraint on the neutrino interaction modeling and oscillation parameter constraints.
This project uses a Convolutional Visual Network to identify the true vertex of NOvA neutrino interactions. We use NOvA's pixel maps (or 'cvnmaps') as the features and the true vertex location as the labels.
In this project, each vertex coordinate (x, y, z) is trained separately. So we have three models for each coordinate. The training for each coordinate is done with the xz and yz pixel maps of the detector.
The code is divided by each {production, detector} sample. More information can be found within each directory.
This project uses:
python 3.7.4tensorflow 2.3.1.