Quanvolutional Neural Networks for Spectrum Peak-Finding

TL;DR

This paper investigates the use of quantum-inspired convolutional neural networks to improve peak detection and quantification in spectral analysis, demonstrating superior performance over classical CNNs on synthetic NMR data.

Contribution

It introduces a simple Quanvolutional Neural Network architecture for spectrum peak-finding, showing its advantages over classical CNNs in accuracy and stability.

Findings

QuanvNNs outperform CNNs with 11% higher F1 score.

QuanvNNs reduce mean absolute error by 30% in peak position estimation.

QuanvNNs show better convergence stability on complex spectra.

Abstract

The analysis of spectra, such as Nuclear Magnetic Resonance (NMR) spectra, for the comprehensive characterization of peaks is a challenging task for both experts and machines, especially with complex molecules. This process, also known as deconvolution, involves identifying and quantifying the peaks in the spectrum. Machine learning techniques have shown promising results in automating this process. With the advent of quantum computing, there is potential to further enhance these techniques. In this work, inspired by the success of classical Convolutional Neural Networks (CNNs), we explore the use of Quanvolutional Neural Networks (QuanvNNs) for the multi-task peak finding problem, involving both peak counting and position estimation. We implement a simple and interpretable QuanvNN architecture that can be directly compared to its classical CNN counterpart, and evaluate its performance on a synthetic NMR-inspired dataset. Our results demonstrate that QuanvNNs outperform classical CNNs on challenging spectra, achieving an 11\% improvement in F1 score and a 30\% reduction in mean absolute error for peak position estimation. Additionally, QuanvNNs appear to exhibit better convergence stability for harder problems.