Emerging generalization advantage of quantum-inspired machine learning in the diagnosis of hepatocellular carcinoma

Aug 4, 2026

Upload by Skyler

Quantum advantage research is increasingly interdisciplinary, with quantum machine learning offering strong generalization potential. This work applies quantum-inspired learning to classify hepatocellular carcinoma tissue using microarray gene expression data. By leveraging previously characterized genetic communities, the approach reduces computational complexity tied to qubit count, enabling quantum-inspired algorithms on classical hardware. Two algorithm families are studied: parameterized quantum circuits (PQC) and tensor networks, comparing accuracy and parameter efficiency on independent test data.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages19

Reading Time48 min

AccessFree to download

Abstract

Article highlights

Introduction

Research Emerging generalization advantage of quantum‑inspired machine learning in the diagnosis of hepatocellular carcinoma Domenico Pomarico1,2 · Alfonso Monaco1,2 · Nicola Amoroso2,3 · Loredana Bellantuono2,4 · Antonio Lacalamita1,2 · Marianna La Rocca1,2 · Tommaso Maggipinto1,2 · Ester Pantaleo1,2 · Sabina Tangaro2,5 · Sebastiano Stramaglia1,2 · Roberto Bellotti1,2 Received: 6 June 2024 / Accepted: 21 February 2025 © The Author(s) 2025 OPEN Abstract Research into quantum advantage is increasingly taking on an interdisciplinary character. In particular, quantum machine learning shows promising generalization capabilities, which we have exploited in the classification of hepatocellular carcinoma tissue based on microarray gene expressions. By using previously characterized genetic communities, we minimize the computational complexity associated with the number of qubits, enabling the execution of quantum-inspired algorithms on classical machines. We consider two categories of such algorithms: parameterized quantum circuits (PQC) and tensor networks. The variational optimization of PQCs achieves better accuracy than classical counterparts on the independent test set, reaching an advantage equal to in accuracy, while tensor networks offer equivalent performance with fewer parameters. Article highlights • We applied QML methods in the classification of hepatocellular carcinoma tissue based on microarray data. • Quantum circuits shows higher performance on independent sets compared to classical methods. • Tensor network methods can match the performance of artificial neural networks (ANNs) while using fewer parameters. Keywords Quantum machine learning · Genes community · Generalization Sebastiano Stramaglia and Roberto Bellotti have contributed equally to this work. * Alfonso Monaco, [alfonso.monaco@ba.infn.it](alfonso.monaco@ba.infn.it); Domenico Pomarico, [domenico.pomarico@ba.infn.it](domenico.pomarico@ba.infn.it); Nicola Amoroso, nicola.amoroso@ [uniba.it](uniba.it); Loredana Bellantuono, [loredana.bellantuono@ba.infn.it](loredana.bellantuono@ba.infn.it); Antonio Lacalamita, antonio. lacalamita@ba. infn. it; Marianna La Rocca, [marianna.larocca@uniba.it](marianna.larocca@uniba.it); Tommaso Maggipinto, [tommaso.maggipinto@ba.infn.it](tommaso.maggipinto@ba.infn.it); Ester Pantaleo, [ester.pantaleo@uniba.it](ester.pantaleo@uniba.it); Sabina Tangaro, [sonia.tangaro@ba.infn.it](sonia.tangaro@ba.infn.it); Sebastiano Stramaglia, [sebastiano.stramaglia@ba.infn.it](sebastiano.stramaglia@ba.infn.it); Roberto Bellotti, [roberto.bellotti@ba.infn.it](roberto.bellotti@ba.infn.it) [|](|1Dipartimento Interateneo di Fisica)[1](|1Dipartimento Interateneo di Fisica)[Dipartimento Interateneo di Fisica](|1Dipartimento Interateneo di Fisica), [Universit](Universit)à [degli Studi di Bari](degli Studi di Bari), [70125 Bari](70125 Bari), [Italy](Italy). 2Istituto Nazionale di Fisica Nucleare, Sezione di Bari, 70125 Bari, Italy. 3Dipartimento di Farmacia-Scienze del Farmaco, Università degli Studi di Bari, 70125 Bari, Italy. 4Dipartimento di BiomedicinaTraslazionale e Neuroscienze (DiBraiN), Università degli Studi di Bari, 70124 Bari, Italy. 5Dipartimento Di Scienze Del Suolo, Della Pianta e Degli Alimenti, Università degli Studi di Bari, 70125 Bari, Italy. Discover Applied Sciences (2025) 7:205 | [https://doi.org/10.1007/s42452-025-06638-6](https://doi.org/10.1007/s42452-025-06638-6) 1 Introduction The constantly growing availability of data and the increasing complexity in analyzing them are driving the development of new technologies that leverage quantum computing to achieve a more efficient integration of data with respect to classical approaches. Quantum applications support these efforts in two ways: by exploiting parallel quantum state superposition [1] and by estimating unknown parameters with unparalleled efficiency [2, 3] . The literature presents a plethora of quantum machine learning (QML) a

FAQs

What problem does the document address?

It investigates whether quantum-inspired machine learning can generalize effectively for classifying hepatocellular carcinoma tissue based on microarray gene expression data.

How do the authors enable quantum-inspired algorithms to run on classical machines?

They use previously characterized genetic communities to minimize computational complexity related to qubit count, making it possible to execute quantum-inspired methods on classical hardware.

Which two types of quantum-inspired algorithms are compared?

The document compares parameterized quantum circuits (PQC) and tensor networks, evaluating their accuracy and efficiency on an independent test set.

Aug 4, 2026

Upload by Skyler