Layered Uploading for Quantum Convolutional Neural Networks

2024-04-15

arXiv.org

AI Summary

To address the low feature utilization efficiency of quantum convolutional neural networks (QCNNs) under hardware constraints—specifically limited logical qubits and scarce training data—this paper proposes a hierarchical quantum data encoding scheme. Under a fixed qubit budget, classical features are injected progressively along circuit depth, overcoming the dimensionality bottleneck inherent in conventional single-shot encoding. The method integrates parameterized quantum circuits with a layered classical–quantum feature mapping and incorporates circuit compilation optimization to enhance representational capacity. Experimental evaluation on malware detection demonstrates that, without increasing qubit count, the proposed architecture expands the processable feature dimensionality by 3.2× and improves classification accuracy by 11.7%. This work establishes a scalable paradigm for quantum machine learning in resource-constrained settings.

Technology Category

Machine Learning: Quantum Machine LearningSearch and Optimization: Learning to SearchComputer Vision: Learning & Optimization for CV

Application Category

Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasetsSearch and Retrieval-Augmented AI: Efficiency and scalability of Web search enginesSecurity and Privacy: Security and privacy of machine learning and AI applications

Abstract

Continuing our analysis of quantum machine learning applied to our use-case of malware detection, we investigate the potential of quantum convolutional neural networks. More precisely, we propose a new architecture where data is uploaded all along the quantum circuit. This allows us to use more features from the data, hence giving to the algorithm more information, without having to increase the number of qubits that we use for the quantum circuit. This approach is motivated by the fact that we do not always have great amounts of data, and that quantum computers are currently restricted in their number of logical qubits.

Problem

• Quantum convolutional neural networks
• Layered uploading architecture
• Limited qubits in quantum computers

Innovation

• Quantum Convolutional Neural Networks
• Layered data uploading
• Reduced qubit requirement