Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Log inSign up
Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Blog
Job Board
Q3AS

© 2026 Aqora Quantum S.A.S.

TermsPrivacyLegal Notice

Neurocomputing

Quantum clustering using kernel entropy component analysis

Yangyang Li, Yang Wang, Yuying Wang, Licheng Jiao, Yang Liu

Published Apr 13, 2016• 24 citations
Publisher
DOI
10.1016/j.neucom.2016.03.006

Related papers

Quantum Kernel Functions for the Prognosis and Health Management of Ball-Bearing Elements

Gabriel San Martín Silva, Enrique López Droguett

This paper showcases the application of quantum-based kernel methods in Prognosis and Health Management tasks for rotating machinery. We provide a concise yet complete introduction to quantum computing from a probabilistic perspective, describing how a quantum algorithm can be interpreted as a kernel function. Using this framework, we assess the efficacy of quantum kernels in two relevant machine learning domains: the categorization of unlabeled data through the application of a quantum-based clustering algorithm, and the reduction of high dimensional data through quantum-based kernel principal component analysis. These approaches are evaluated using a well-known benchmark dataset of ball-bearing elements. The dataset, originally in the form of time series data collected from accelerometer sensors, consists of three distinct classes, and is first processed to extract pertinent time features. Results indicate that the proposed quantum-based approach performs comparably to traditional methodologies, motivating future research efforts in the use of Quantum Kernels Functions for machine learning applications in the field of Prognosis and Health Management.

IEEE Conference on Prognostics and Health Management•Jun 17, 2024•1 citation

Clustering Structure and Quantum Computing

Péter Wittek

Elsevier eBooks•Jan 1, 2014•6 citations

Beyond Euclidean Clustering: Quantum Kernel-Driven Decomposition for the Capacitated Vehicle Routing Problem

Jun Towada

Vehicle routing problems (VRPs) remain challenging at scale under practical time budgets. Since noisy intermediate-scale quantum (NISQ) devices are not yet suitable for end-to-end capacitated VRP (CVRP) optimization, we use quantum processors as feature extractors for structural preprocessing. We propose a hybrid quantum-classical pipeline in which a shallow quantum feature map induces a quantum kernel that defines nonlinear customer-to-customer similarities. The induced similarities drive a spectral-clustering-based decomposition, which is refined by capacity adjustment, and the resulting subproblems are solved by a classical routing pipeline with RL-based initialization and adaptive large neighborhood search (ALNS). Compared with a strong classical baseline that applies K-means on the same feature space, we find that higher geometric cluster separability (e.g., Silhouette score) does not necessarily yield better downstream routing. In contrast, quantum-kernel-driven decomposition achieves smaller routing optimality gaps (sub-2% under our evaluation setting) on difficult instances such as CMT1, indicating that quantum kernels can capture latent routing-aware structure beyond Euclidean clustering. Experiments on IBM's ibm_kobe device further show that clustering-relevant kernel structure remains observable under hardware noise, supporting the feasibility of NISQ-era quantum kernels as scalable structural components in logistics optimization.

•Feb 20, 2026•0 citations

Quantum Clustering

Davide Pastorello

Machine learning•Dec 16, 2022•1 citation

Quantum Kernels for Narrative Coherence: An Application to Path Optimization in Document Graphs for Storyline Extraction

Brian Felipe Keith-Norambuena, Javiera Canales, Maximiliano Araya +4 more

Narrative extraction algorithms construct storylines by finding coherent paths through document collections. The Narrative Trails algorithm frames this as maximum-capacity path optimization, where path quality depends on a coherence function measuring document relationships. We introduce quantum kernels as coherence functions for narrative extraction—to the best of our knowledge, the first systematic characterisation of quantum kernel methods for storyline extraction—and compare them against classical baselines on two corpora using a multi-seed protocol. The sweep covers 93 method evaluations (54 quantum kernels across three encoder families—RY+CNOT-ring, IQP/ZZ-feature-map, and a projected quantum kernel—and 39 classical kernels—cosine, RBF, and the cluster-aware Narrative Trails baseline). On 11,215 human navigation paths from Wikispeedia, evaluation metrics divide into two clusters that disagree with each other: alignment-based metrics (length-normalised DTW and per-step DTW similarity) favour methods that produce long alignment-rich paths, while set-overlap metrics (Jaccard and F1) favour methods that produce shorter paths with higher article overlap. On LLM-judged coherence for Cuban news storylines, evaluated under a 12-method × 5-seed × 30-endpoint-pair × 2-judge design (Claude Sonnet 4.5 and GPT-4o, both at T=0 via structured tool calling), the cluster-aware classical baseline is the top method in terms of mean overall coherence; the 5-method quantum-kernel pool and the 7-method classical-kernel pool on matched projection input show no significant differences after Holm correction. Cross-task analysis reveals that LLM coherence rank correlates with alignment-cluster Wikispeedia metrics (Spearman ρ≈+0.70) and anti-correlates with overlap-cluster metrics (ρ≈−0.62). A closed-form theoretical analysis shows that the depth-1 RY+CNOT-ring kernel reduces to a classical product-of-cosines kernel order equivalent to RBF, explaining the absence of empirical separation at low depth; deeper encoders break the cancellation but exponentially concentrate kernel values, eroding inter-pair distinguishability. Our results characterise quantum coherence kernels as competitive with classical kernels on the same projected input rather than decisively superior, with the cluster-aware classical baseline retaining a modest advantage attributable to its explicit topical structure.

Mathematics•May 18, 2026•0 citations

Clustering Within Quantum Mechanical Framework

Güleser Kalaycı Demir

Lecture notes in computer science•Jan 1, 2005•2 citations
This paper showcases the application of quantum-based kernel methods in Prognosis and Health Management tasks for rotating machinery. We provide a concise yet complete introduction to quantum computing from a probabilistic perspective, describing how a quantum algorithm can be interpreted as a kernel function. Using this framework, we assess the efficacy of quantum kernels in two relevant machine learning domains: the categorization of unlabeled data through the application of a quantum-based clustering algorithm, and the reduction of high dimensional data through quantum-based kernel principal component analysis. These approaches are evaluated using a well-known benchmark dataset of ball-bearing elements. The dataset, originally in the form of time series data collected from accelerometer sensors, consists of three distinct classes, and is first processed to extract pertinent time features. Results indicate that the proposed quantum-based approach performs comparably to traditional methodologies, motivating future research efforts in the use of Quantum Kernels Functions for machine learning applications in the field of Prognosis and Health Management.
Vehicle routing problems (VRPs) remain challenging at scale under practical time budgets. Since noisy intermediate-scale quantum (NISQ) devices are not yet suitable for end-to-end capacitated VRP (CVRP) optimization, we use quantum processors as feature extractors for structural preprocessing. We propose a hybrid quantum-classical pipeline in which a shallow quantum feature map induces a quantum kernel that defines nonlinear customer-to-customer similarities. The induced similarities drive a spectral-clustering-based decomposition, which is refined by capacity adjustment, and the resulting subproblems are solved by a classical routing pipeline with RL-based initialization and adaptive large neighborhood search (ALNS). Compared with a strong classical baseline that applies K-means on the same feature space, we find that higher geometric cluster separability (e.g., Silhouette score) does not necessarily yield better downstream routing. In contrast, quantum-kernel-driven decomposition achieves smaller routing optimality gaps (sub-2% under our evaluation setting) on difficult instances such as CMT1, indicating that quantum kernels can capture latent routing-aware structure beyond Euclidean clustering. Experiments on IBM's ibm_kobe device further show that clustering-relevant kernel structure remains observable under hardware noise, supporting the feasibility of NISQ-era quantum kernels as scalable structural components in logistics optimization.
Narrative extraction algorithms construct storylines by finding coherent paths through document collections. The Narrative Trails algorithm frames this as maximum-capacity path optimization, where path quality depends on a coherence function measuring document relationships. We introduce quantum kernels as coherence functions for narrative extraction—to the best of our knowledge, the first systematic characterisation of quantum kernel methods for storyline extraction—and compare them against classical baselines on two corpora using a multi-seed protocol. The sweep covers 93 method evaluations (54 quantum kernels across three encoder families—RY+CNOT-ring, IQP/ZZ-feature-map, and a projected quantum kernel—and 39 classical kernels—cosine, RBF, and the cluster-aware Narrative Trails baseline). On 11,215 human navigation paths from Wikispeedia, evaluation metrics divide into two clusters that disagree with each other: alignment-based metrics (length-normalised DTW and per-step DTW similarity) favour methods that produce long alignment-rich paths, while set-overlap metrics (Jaccard and F1) favour methods that produce shorter paths with higher article overlap. On LLM-judged coherence for Cuban news storylines, evaluated under a 12-method × 5-seed × 30-endpoint-pair × 2-judge design (Claude Sonnet 4.5 and GPT-4o, both at T=0 via structured tool calling), the cluster-aware classical baseline is the top method in terms of mean overall coherence; the 5-method quantum-kernel pool and the 7-method classical-kernel pool on matched projection input show no significant differences after Holm correction. Cross-task analysis reveals that LLM coherence rank correlates with alignment-cluster Wikispeedia metrics (Spearman ρ≈+0.70) and anti-correlates with overlap-cluster metrics (ρ≈−0.62). A closed-form theoretical analysis shows that the depth-1 RY+CNOT-ring kernel reduces to a classical product-of-cosines kernel order equivalent to RBF, explaining the absence of empirical separation at low depth; deeper encoders break the cancellation but exponentially concentrate kernel values, eroding inter-pair distinguishability. Our results characterise quantum coherence kernels as competitive with classical kernels on the same projected input rather than decisively superior, with the cluster-aware classical baseline retaining a modest advantage attributable to its explicit topical structure.