Algorithmic Strategies for seizing Quantum Computing

TL;DR

This paper explores two quantum algorithm strategies, re-uploading and unary, to leverage quantum properties for improved performance and noise resilience on current noisy quantum computers.

Contribution

It introduces the re-uploading and unary strategies as novel methods to enhance quantum algorithm robustness and effectiveness on noisy hardware.

Findings

Re-uploading enables quantum circuits to learn complex behaviors.

Unary strategy reduces information density to improve noise resilience.

Small quantum speed-ups are achievable with unary in noisy devices.

Abstract

Quantum computing is a nascent technology with prospects to have a huge impact in the world. Its current status, however, only counts on small and noisy quantum computers whose performance is limited. In this thesis, two different strategies are explored to take advantage of inherently quantum properties and propose recipes to seize quantum computing since its advent. First, the re-uploading strategy is a variational algorithm related to machine learning. It consists in introducing data several times along a computation accompanied by tunable parameters. This process permits the circuit to learn and mimic any behavior. This capability emerges naturally from the quantum properties of the circuit. Second, the unary strategy aims to reduce the density of information stored in a quantum circuit to increase its resilience against noise. This trade-off between performance and robustness brings an advantage for noisy devices, where small but meaningful quantum speed-ups can be found.