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

arXiv:2605.03565 · IEEE Annual International Computer Software and Applications Conference (COMPSAC)

Neural optimization for quantum architectures: graph embedding problems with Distance Encoder Networks

Chiara Vercellino, Giacomo Vitali, Paolo Viviani, Alberto Scionti, Andrea Scarabosio, Olivier Terzo, Edoardo Giusto, Bartolomeo Montrucchio

quant-phSubmitted May 5, 2026• Announced May 6, 2026• 3 citations
arXivHTMLPDF

Abstract

Quantum machines are among the most promising technologies expected to provide significant improvements in the following years. However, bridging the gap between real-world applications and their implementation on quantum hardware is still a complicated task. One of the main challenges is to represent through qubits (i.e., the basic units of quantum information) the problems of interest. According to the specific technology underlying the quantum machine, it is necessary to implement a proper representation strategy, generally referred to as embedding. This paper introduces a neural-enhanced optimization framework to solve the constrained unit disk problem, which arises in the context of qubits positioning for neutral atoms-based quantum hardware. The proposed approach involves a modified autoencoder model, i.e., the Distances Encoder Network, and a custom loss, i.e., the Embedding Loss Function, respectively, to compute Euclidean distances and model the optimization constraints. The core idea behind this design relies on the capability of neural networks to approximate non-linear transformations to make the Distances Encoder Network learn the spatial transformation that maps initial non-feasible solutions of the constrained unit disk problem into feasible ones. The proposed approach outperforms classical solvers, given fixed comparable computation times, and paves the way to address other optimization problems through a similar strategy.

Journal reference
2023 IEEE 47th Annual Computers, Software, and Applications Conference (COMPSAC). IEEE, 2023. p. 380-389
DOI
10.1109/COMPSAC57700.2023.00058

Related papers

Hilbert-Space Quantum Neural Representation Learning for NISQ-Era Quantum Computing

Yijie Zhu

The convergence of quantum computing and machine learning offers significant potential for processing complex data distributions. However, the realization of this potential in the Noisy Intermediate-Scale Quantum (NISQ) era is constrained by physical hardware limitations, particularly restricted qubit counts and system noise. This thesis addresses the challenge of the ”Representation Bottleneck” by establishing a framework for resource-efficient quantum representation learning. The research demonstrates that effective learning on near-term devices can be achieved by optimizing circuit architectures and objective functions to align with hardware constraints. First, to address resource limitations in high-dimensional visual recognition, we propose the Multi-Gate Quantum Convolutional Neural Network (MG-QNN). By employing a Multi-Gate strategy, this architecture optimizes the utilization of quan tum resources within the circuit design. Empirical evaluations on face recognition benchmarks indicate that this approach maintains competitive recognition accuracy while reducing the qubit requirement compared to standard quantum convolutional methods. Second, extending the framework to non-Euclidean data, we introduce the Single-Qubit Graph Neural Network (sQGNN). Utilizing Discrete-Time Quantum Walks, this method encodes graph topologies and relational patterns onto a single qubit. This approach circumvents the need for multi-qubit entangling gates, thereby reducing the computational overhead and noise sensitivity typically associated with processing graph-structured data. Third, in the domain of generative modeling, we present the Single-Qubit Quan tum Convolutional Variational Autoencoder (QCVAE). Addressing the difficulty of generative tasks on constrained devices, this model integrates the Information Bottleneck principle with efficient data re-uploading techniques. The study demonstrates that a single qubit possesses sufficient expressivity to model and reconstruct image distributions effectively when guided by appropriate information theoretic constraints. Finally, to resolve the fundamental limitations of classical loss functions in quantum learning, we establish the Hilbert-Schmidt Information Bottleneck (HSIB). We identify that standard Euclidean metrics often fail to respect the geometric structure of the Hilbert space. The HSIB framework employs the Hilbert Schmidt distance as a tractable geometric measure, providing a learning objective that maximizes information retention while adhering to the intrinsic geometry of quantum states. Collectively, these contributions illustrate that practical quantum machine learning advantage in the NISQ era does not rely solely on scaling hardware, but on the strategic design of resource-efficient architectures and geometrically valid metrics.

•Aug 17, 2026•0 citations

Towards Quantum Graph Neural Networks: An Ego-Graph Learning Approach

Ai, Xing, Zhihong Zhang, Luzhe Sun +2 more

Quantum machine learning is a fast-emerging field that aims to tackle machine learning using quantum algorithms and quantum computing. Due to the lack of physical qubits and an effective means to map real-world data from Euclidean space to Hilbert space, most of these methods focus on quantum analogies or process simulations rather than devising concrete architectures based on qubits. In this paper, we propose a novel hybrid quantum-classical algorithm for graph-structured data, which we refer to as the Ego-graph based Quantum Graph Neural Network (egoQGNN). egoQGNN implements the GNN theoretical framework using the tensor product and unity matrix representation, which greatly reduces the number of model parameters required. When controlled by a classical computer, egoQGNN can accommodate arbitrarily sized graphs by processing ego-graphs from the input graph using a modestly-sized quantum device. The architecture is based on a novel mapping from real-world data to Hilbert space. This mapping maintains the distance relations present in the data and reduces information loss. Experimental results show that the proposed method outperforms competitive state-of-the-art models with only 1.68\% parameters compared to those models.

arXiv:2201.05158•Jan 13, 2022•32 citations

Interaction-Aware Embedding Optimization for Neutral-Atom Quantum Processors

Clément de Terrasson de Montleau, Victor Drouin-Touchette, Wesley Coelho

Neutral atom-based quantum computers have the ability to dynamically reshape the qubit register geometry between experimental runs. This capability enables different types of problems to be mapped directly onto the hardware, significantly enhancing the efficiency of solving, for instance, complex optimization and machine learning use-cases. However, this flexibility requires finding atom positions to implement target pairwise interactions, inducing a difficult nonconvex optimization for setting up the quantum machine. In this work, we introduce a new force-directed method that targets interactions to encode the input problem on the quantum computer. Numerical results show that our new method can improve embedding feasibility and solution quality by orders of magnitude, and remains reliable on the largest instances.

quant-pharXiv:2610.04068•Oct 6, 2026•2 citations

Quantum Annealing for Combinatorial Optimization with Graph Neural Networks

Jincheng Zhang

This paper investigates the application of quantum annealing in conjunction with graph neural networks (GNNs) for solving complex combinatorial optimization problems. Traditional classical optimization algorithms often face significant challenges in identifying optimal solutions within large and intricate solution spaces. The core idea presented herein is to represent these problems as graphs and utilize the inherent quantum mechanical properties of quantum annealing to explore this space more efficiently. Specifically, GNNs are employed to generate embeddings of the graph, effectively encoding problem-specific features and relationships. This approach aims to bridge the gap between the probabilistic nature of quantum computation and the structural information captured by GNNs. We detail the theoretical framework, explore the potential benefits of this hybrid approach, and outline a methodology for integrating these techniques. The potential for improved optimization performance compared to conventional methods is discussed, highlighting the synergy between quantum annealing's ability to find low-energy states and GNNs' capacity to learn complex graph representations.

•Aug 31, 2026•0 citations

Quantum State Adaptative Encoding and Encoder Design

Jincheng Zhang

Quantum State Adaptative Encoding is a novel approach to quantum computing that utilizes machine learning to dynamically generate optimal encodings for quantum states. Traditional encoding methods often rely on pre-defined encodings, limiting the adaptability of quantum algorithms. This research proposes a system that learns the underlying characteristics of quantum states and automatically designs encodings tailored to specific computational tasks. We present a framework based on a neural network to capture and optimize the quantum state, resulting in enhanced efficiency and accuracy in quantum computations. The core mechanism involves a self-learning process where the network adjusts the encoding to maximize the quantum state's utility. This addresses the limitations of fixed encoding schemes and paves the way for more efficient and versatile quantum algorithms.

•Aug 31, 2026•0 citations

Self-Organizing Quantum Neural Networks for Optimal Resource Allocation

Jincheng Zhang

This paper investigates the application of self-organizing quantum neural networks (SQNNs) to the problem of optimal resource allocation, a core challenge across diverse domains including finance, logistics, and telecommunications. Traditional classical neural networks frequently struggle to achieve optimal solutions when faced with complex constraints and dynamic environments. The central hypothesis of this work is that leveraging quantum computation through SQNNs, specifically utilizing qubit representations of resource states and exploiting entanglement, can significantly enhance the exploration of the solution space and lead to near-optimal resource allocation decisions. We propose a novel SQNN architecture designed to address this challenge. The architecture incorporates layers of quantum neurons that dynamically adjust their connectivity based on the input data, mimicking the self-organizing properties observed in biological neural networks. We present a theoretical framework outlining the key components of the SQNN and detail the mathematical formulation of the resource allocation problem being addressed. The core of the model is expressed as follows: Let *N* be the number of resources to be allocated. Let *K* be the number of potential allocation options for each resource. Let *xi* represent the allocation decision for resource *i*, where *xi* ∈ {1, 2, ..., *K*}. The objective is to minimize the cost function *C(x)*, which is defined as: *C(x) = ∑i=1N *Ci(xi)* where *Ci(xi)* represents the cost associated with resource *i* given the allocation *xi*. This cost function is subject to constraints *G(x) ≤ 0*, where *G* is a vector of constraints representing factors like capacity limits, demand requirements, and regulatory restrictions. The SQNN learns to approximate the optimal solution *x* that minimizes *C(x)* subject to the constraints *G(x) ≤ 0*. The network's self-organizing capability is achieved through a modified backpropagation algorithm adapted for quantum systems, incorporating quantum measurement and entangling operations to refine the network's internal state and improve its ability to represent and optimize the resource allocation problem. The efficacy of the SQNN is evaluated through simulations, demonstrating its potential to outperform classical neural networks in complex resource allocation scenarios.

•Sep 2, 2026•0 citations
Quantum machines are among the most promising technologies expected to provide significant improvements in the following years. However, bridging the gap between real-world applications and their implementation on quantum hardware is still a complicated task. One of the main challenges is to represent through qubits (i.e., the basic units of quantum information) the problems of interest. According to the specific technology underlying the quantum machine, it is necessary to implement a proper representation strategy, generally referred to as embedding. This paper introduces a neural-enhanced optimization framework to solve the constrained unit disk problem, which arises in the context of qubits positioning for neutral atoms-based quantum hardware. The proposed approach involves a modified autoencoder model, i.e., the Distances Encoder Network, and a custom loss, i.e., the Embedding Loss Function, respectively, to compute Euclidean distances and model the optimization constraints. The core idea behind this design relies on the capability of neural networks to approximate non-linear transformations to make the Distances Encoder Network learn the spatial transformation that maps initial non-feasible solutions of the constrained unit disk problem into feasible ones. The proposed approach outperforms classical solvers, given fixed comparable computation times, and paves the way to address other optimization problems through a similar strategy.
The convergence of quantum computing and machine learning offers significant potential for processing complex data distributions. However, the realization of this potential in the Noisy Intermediate-Scale Quantum (NISQ) era is constrained by physical hardware limitations, particularly restricted qubit counts and system noise. This thesis addresses the challenge of the ”Representation Bottleneck” by establishing a framework for resource-efficient quantum representation learning. The research demonstrates that effective learning on near-term devices can be achieved by optimizing circuit architectures and objective functions to align with hardware constraints. First, to address resource limitations in high-dimensional visual recognition, we propose the Multi-Gate Quantum Convolutional Neural Network (MG-QNN). By employing a Multi-Gate strategy, this architecture optimizes the utilization of quan tum resources within the circuit design. Empirical evaluations on face recognition benchmarks indicate that this approach maintains competitive recognition accuracy while reducing the qubit requirement compared to standard quantum convolutional methods. Second, extending the framework to non-Euclidean data, we introduce the Single-Qubit Graph Neural Network (sQGNN). Utilizing Discrete-Time Quantum Walks, this method encodes graph topologies and relational patterns onto a single qubit. This approach circumvents the need for multi-qubit entangling gates, thereby reducing the computational overhead and noise sensitivity typically associated with processing graph-structured data. Third, in the domain of generative modeling, we present the Single-Qubit Quan tum Convolutional Variational Autoencoder (QCVAE). Addressing the difficulty of generative tasks on constrained devices, this model integrates the Information Bottleneck principle with efficient data re-uploading techniques. The study demonstrates that a single qubit possesses sufficient expressivity to model and reconstruct image distributions effectively when guided by appropriate information theoretic constraints. Finally, to resolve the fundamental limitations of classical loss functions in quantum learning, we establish the Hilbert-Schmidt Information Bottleneck (HSIB). We identify that standard Euclidean metrics often fail to respect the geometric structure of the Hilbert space. The HSIB framework employs the Hilbert Schmidt distance as a tractable geometric measure, providing a learning objective that maximizes information retention while adhering to the intrinsic geometry of quantum states. Collectively, these contributions illustrate that practical quantum machine learning advantage in the NISQ era does not rely solely on scaling hardware, but on the strategic design of resource-efficient architectures and geometrically valid metrics.
Quantum machine learning is a fast-emerging field that aims to tackle machine learning using quantum algorithms and quantum computing. Due to the lack of physical qubits and an effective means to map real-world data from Euclidean space to Hilbert space, most of these methods focus on quantum analogies or process simulations rather than devising concrete architectures based on qubits. In this paper, we propose a novel hybrid quantum-classical algorithm for graph-structured data, which we refer to as the Ego-graph based Quantum Graph Neural Network (egoQGNN). egoQGNN implements the GNN theoretical framework using the tensor product and unity matrix representation, which greatly reduces the number of model parameters required. When controlled by a classical computer, egoQGNN can accommodate arbitrarily sized graphs by processing ego-graphs from the input graph using a modestly-sized quantum device. The architecture is based on a novel mapping from real-world data to Hilbert space. This mapping maintains the distance relations present in the data and reduces information loss. Experimental results show that the proposed method outperforms competitive state-of-the-art models with only 1.68\% parameters compared to those models.
Neutral atom-based quantum computers have the ability to dynamically reshape the qubit register geometry between experimental runs. This capability enables different types of problems to be mapped directly onto the hardware, significantly enhancing the efficiency of solving, for instance, complex optimization and machine learning use-cases. However, this flexibility requires finding atom positions to implement target pairwise interactions, inducing a difficult nonconvex optimization for setting up the quantum machine. In this work, we introduce a new force-directed method that targets interactions to encode the input problem on the quantum computer. Numerical results show that our new method can improve embedding feasibility and solution quality by orders of magnitude, and remains reliable on the largest instances.
This paper investigates the application of quantum annealing in conjunction with graph neural networks (GNNs) for solving complex combinatorial optimization problems. Traditional classical optimization algorithms often face significant challenges in identifying optimal solutions within large and intricate solution spaces. The core idea presented herein is to represent these problems as graphs and utilize the inherent quantum mechanical properties of quantum annealing to explore this space more efficiently. Specifically, GNNs are employed to generate embeddings of the graph, effectively encoding problem-specific features and relationships. This approach aims to bridge the gap between the probabilistic nature of quantum computation and the structural information captured by GNNs. We detail the theoretical framework, explore the potential benefits of this hybrid approach, and outline a methodology for integrating these techniques. The potential for improved optimization performance compared to conventional methods is discussed, highlighting the synergy between quantum annealing's ability to find low-energy states and GNNs' capacity to learn complex graph representations.
Quantum State Adaptative Encoding is a novel approach to quantum computing that utilizes machine learning to dynamically generate optimal encodings for quantum states. Traditional encoding methods often rely on pre-defined encodings, limiting the adaptability of quantum algorithms. This research proposes a system that learns the underlying characteristics of quantum states and automatically designs encodings tailored to specific computational tasks. We present a framework based on a neural network to capture and optimize the quantum state, resulting in enhanced efficiency and accuracy in quantum computations. The core mechanism involves a self-learning process where the network adjusts the encoding to maximize the quantum state's utility. This addresses the limitations of fixed encoding schemes and paves the way for more efficient and versatile quantum algorithms.
This paper investigates the application of self-organizing quantum neural networks (SQNNs) to the problem of optimal resource allocation, a core challenge across diverse domains including finance, logistics, and telecommunications. Traditional classical neural networks frequently struggle to achieve optimal solutions when faced with complex constraints and dynamic environments. The central hypothesis of this work is that leveraging quantum computation through SQNNs, specifically utilizing qubit representations of resource states and exploiting entanglement, can significantly enhance the exploration of the solution space and lead to near-optimal resource allocation decisions. We propose a novel SQNN architecture designed to address this challenge. The architecture incorporates layers of quantum neurons that dynamically adjust their connectivity based on the input data, mimicking the self-organizing properties observed in biological neural networks. We present a theoretical framework outlining the key components of the SQNN and detail the mathematical formulation of the resource allocation problem being addressed. The core of the model is expressed as follows: Let *N* be the number of resources to be allocated. Let *K* be the number of potential allocation options for each resource. Let *xi* represent the allocation decision for resource *i*, where *xi* ∈ {1, 2, ..., *K*}. The objective is to minimize the cost function *C(x)*, which is defined as: *C(x) = ∑i=1N *Ci(xi)* where *Ci(xi)* represents the cost associated with resource *i* given the allocation *xi*. This cost function is subject to constraints *G(x) ≤ 0*, where *G* is a vector of constraints representing factors like capacity limits, demand requirements, and regulatory restrictions. The SQNN learns to approximate the optimal solution *x* that minimizes *C(x)* subject to the constraints *G(x) ≤ 0*. The network's self-organizing capability is achieved through a modified backpropagation algorithm adapted for quantum systems, incorporating quantum measurement and entangling operations to refine the network's internal state and improve its ability to represent and optimize the resource allocation problem. The efficacy of the SQNN is evaluated through simulations, demonstrating its potential to outperform classical neural networks in complex resource allocation scenarios.