On Quantum Perceptron Learning via Quantum Search

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Sun, Xiaoyu, Roget, Mathieu, Di Molfetta, Giuseppe, Kadri, Hachem
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866909547528978432
author Sun, Xiaoyu
Roget, Mathieu
Di Molfetta, Giuseppe
Kadri, Hachem
author_facet Sun, Xiaoyu
Roget, Mathieu
Di Molfetta, Giuseppe
Kadri, Hachem
contents With the growing interest in quantum machine learning, the perceptron -- a fundamental building block in traditional machine learning -- has emerged as a valuable model for exploring quantum advantages. Two quantum perceptron algorithms based on Grover's search, were developed in arXiv:1602.04799 to accelerate training and improve statistical efficiency in perceptron learning. This paper points out and corrects a mistake in the proof of Theorem 2 in arXiv:1602.04799. Specifically, we show that the probability of sampling from a normal distribution for a $D$-dimensional hyperplane that perfectly classifies the data scales as $Ω(γ^{D})$ instead of $Θ(γ)$, where $γ$ is the margin. We then revisit two well-established linear programming algorithms -- the ellipsoid method and the cutting plane random walk algorithm -- in the context of perceptron learning, and show how quantum search algorithms can be leveraged to enhance the overall complexity. Specifically, both algorithms gain a sub-linear speed-up $O(\sqrt{N})$ in the number of data points $N$ as a result of Grover's algorithm and an additional $O(D^{1.5})$ speed-up is possible for cutting plane random walk algorithm employing quantum walk search.
format Preprint
id arxiv_https___arxiv_org_abs_2503_17308
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle On Quantum Perceptron Learning via Quantum Search
Sun, Xiaoyu
Roget, Mathieu
Di Molfetta, Giuseppe
Kadri, Hachem
Quantum Physics
Machine Learning
With the growing interest in quantum machine learning, the perceptron -- a fundamental building block in traditional machine learning -- has emerged as a valuable model for exploring quantum advantages. Two quantum perceptron algorithms based on Grover's search, were developed in arXiv:1602.04799 to accelerate training and improve statistical efficiency in perceptron learning. This paper points out and corrects a mistake in the proof of Theorem 2 in arXiv:1602.04799. Specifically, we show that the probability of sampling from a normal distribution for a $D$-dimensional hyperplane that perfectly classifies the data scales as $Ω(γ^{D})$ instead of $Θ(γ)$, where $γ$ is the margin. We then revisit two well-established linear programming algorithms -- the ellipsoid method and the cutting plane random walk algorithm -- in the context of perceptron learning, and show how quantum search algorithms can be leveraged to enhance the overall complexity. Specifically, both algorithms gain a sub-linear speed-up $O(\sqrt{N})$ in the number of data points $N$ as a result of Grover's algorithm and an additional $O(D^{1.5})$ speed-up is possible for cutting plane random walk algorithm employing quantum walk search.
title On Quantum Perceptron Learning via Quantum Search
topic Quantum Physics
Machine Learning
url https://arxiv.org/abs/2503.17308