Accelerating the training of single-layer binary neural networks using the HHL quantum algorithm

TL;DR

This paper explores how the HHL quantum algorithm can be leveraged to accelerate training in single-layer binary neural networks by extracting useful information from quantum solutions to linear regression problems, reducing classical computational complexity.

Contribution

It demonstrates that valuable insights from the HHL quantum algorithm can be used to lower the classical computational burden in training binary neural networks.

Findings

Quantum information from HHL can inform classical solutions.

Reduction in classical complexity for linear regression in neural training.

Potential for more efficient binary neural network training methods.

Abstract

Binary Neural Networks are a promising technique for implementing efficient deep models with reduced storage and computational requirements. The training of these is however, still a compute-intensive problem that grows drastically with the layer size and data input. At the core of this calculation is the linear regression problem. The Harrow-Hassidim-Lloyd (HHL) quantum algorithm has gained relevance thanks to its promise of providing a quantum state containing the solution of a linear system of equations. The solution is encoded in superposition at the output of a quantum circuit. Although this seems to provide the answer to the linear regression problem for the training neural networks, it also comes with multiple, difficult-to-avoid hurdles. This paper shows, however, that useful information can be extracted from the quantum-mechanical implementation of HHL, and used to reduce the complexity of finding the solution on the classical side.