Differentiable Earth Mover's Distance for Data Compression at the High-Luminosity LHC

TL;DR

This paper introduces a neural network-based differentiable approximation of Earth Mover's Distance, enabling efficient training of data compression models for high-energy physics data at CERN.

Contribution

We develop a CNN that approximates EMD differentiably and apply it to train a data compression neural network for particle detector data at the LHC.

Findings

The differentiable EMD CNN is faster and effective for training.

The encoder neural network outperforms MSE-based training.

Improved data compression preserves energy distribution information.

Abstract

The Earth mover's distance (EMD) is a useful metric for image recognition and classification, but its usual implementations are not differentiable or too slow to be used as a loss function for training other algorithms via gradient descent. In this paper, we train a convolutional neural network (CNN) to learn a differentiable, fast approximation of the EMD and demonstrate that it can be used as a substitute for computing-intensive EMD implementations. We apply this differentiable approximation in the training of an autoencoder-inspired neural network (encoder NN) for data compression at the high-luminosity LHC at CERN. The goal of this encoder NN is to compress the data while preserving the information related to the distribution of energy deposits in particle detectors. We demonstrate that the performance of our encoder NN trained using the differentiable EMD CNN surpasses that of training with loss functions based on mean squared error.