Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Neural Architecture Search for Quantum Autoencoders

Hibah Agha, Samuel Yen-Chi Chen, Huan-Hsin Tseng, Shinjae Yoo

TL;DR

This work tackles the problem of designing effective quantum autoencoders for near-term quantum devices by automating circuit design through neural architecture search (NAS) based on genetic algorithms. The authors evolve variational quantum circuit configurations within a hybrid quantum-classical autoencoder framework, coupling topology search with gradient-based parameter optimization. Experiments on MNIST and FashionMNIST demonstrate that mutated, more entangled yet carefully structured circuits can outperform fixed baselines, highlighting the value of automated architectural exploration in quantum machine learning. The findings suggest that evolutionary NAS can yield robust, diverse quantum architectures adaptable to noise and hardware constraints, paving the way for scalable quantum feature learning.

Abstract

In recent years, machine learning and deep learning have driven advances in domains such as image classification, speech recognition, and anomaly detection by leveraging multi-layer neural networks to model complex data. Simultaneously, quantum computing (QC) promises to address classically intractable problems via quantum parallelism, motivating research in quantum machine learning (QML). Among QML techniques, quantum autoencoders show promise for compressing high-dimensional quantum and classical data. However, designing effective quantum circuit architectures for quantum autoencoders remains challenging due to the complexity of selecting gates, arranging circuit layers, and tuning parameters. This paper proposes a neural architecture search (NAS) framework that automates the design of quantum autoencoders using a genetic algorithm (GA). By systematically evolving variational quantum circuit (VQC) configurations, our method seeks to identify high-performing hybrid quantum-classical autoencoders for data reconstruction without becoming trapped in local minima. We demonstrate effectiveness on image datasets, highlighting the potential of quantum autoencoders for efficient feature extraction within a noise-prone, near-term quantum era. Our approach lays a foundation for broader application of genetic algorithms to quantum architecture search, aiming for a robust, automated method that can adapt to varied data and hardware constraints.

Neural Architecture Search for Quantum Autoencoders

TL;DR

This work tackles the problem of designing effective quantum autoencoders for near-term quantum devices by automating circuit design through neural architecture search (NAS) based on genetic algorithms. The authors evolve variational quantum circuit configurations within a hybrid quantum-classical autoencoder framework, coupling topology search with gradient-based parameter optimization. Experiments on MNIST and FashionMNIST demonstrate that mutated, more entangled yet carefully structured circuits can outperform fixed baselines, highlighting the value of automated architectural exploration in quantum machine learning. The findings suggest that evolutionary NAS can yield robust, diverse quantum architectures adaptable to noise and hardware constraints, paving the way for scalable quantum feature learning.

Abstract

In recent years, machine learning and deep learning have driven advances in domains such as image classification, speech recognition, and anomaly detection by leveraging multi-layer neural networks to model complex data. Simultaneously, quantum computing (QC) promises to address classically intractable problems via quantum parallelism, motivating research in quantum machine learning (QML). Among QML techniques, quantum autoencoders show promise for compressing high-dimensional quantum and classical data. However, designing effective quantum circuit architectures for quantum autoencoders remains challenging due to the complexity of selecting gates, arranging circuit layers, and tuning parameters. This paper proposes a neural architecture search (NAS) framework that automates the design of quantum autoencoders using a genetic algorithm (GA). By systematically evolving variational quantum circuit (VQC) configurations, our method seeks to identify high-performing hybrid quantum-classical autoencoders for data reconstruction without becoming trapped in local minima. We demonstrate effectiveness on image datasets, highlighting the potential of quantum autoencoders for efficient feature extraction within a noise-prone, near-term quantum era. Our approach lays a foundation for broader application of genetic algorithms to quantum architecture search, aiming for a robust, automated method that can adapt to varied data and hardware constraints.

Paper Structure

This paper contains 14 sections, 3 equations, 4 figures, 1 algorithm.

Table of Contents

  1. Introduction
  2. Related Work
  3. Quantum Architecture Search
  4. Quantum Autoencoder
  5. Genetic Algorithms
  6. Quantum Autoencoder
  7. Autoencoders
  8. Variational Quantum Circuit
  9. Neural Architecture Search for Quantum Autoencoders
  10. Experiment
  11. MNIST
  12. FashionMNIST
  13. Results
  14. Conclusion

Figures (4)

  • Figure 1: Hybrid Quantum-Classical Computing Paradigm.
  • Figure 2: A VQC diagram. A classical input $x$ is encoded using an operator $V(x)$, followed by the application of the variational circuit $U(\Theta)$. The circuit's output is then measured using a predefined observable.
  • Figure 3: Example of circuit mutations passing down generation by generation.
  • Figure 4: Benchmarking MNIST Data.