Abstract:Quantum embedding is a fundamental prerequisite for applying quantum machine learning techniques to classical data, and has substantial impacts on performance outcomes. In this study, we present Neural Quantum Embedding (NQE), a method that efficiently optimizes quantum embedding beyond the limitations of positive and trace-preserving maps by leveraging classical deep learning techniques. NQE enhances the lower bound of the empirical risk, leading to substantial improvements in classification performance. Moreover, NQE improves robustness against noise. To validate the effectiveness of NQE, we conduct experiments on IBM quantum devices for image data classification, resulting in a remarkable accuracy enhancement from 0.52 to 0.96. In addition, numerical analyses highlight that NQE simultaneously improves the trainability and generalization performance of quantum neural networks, as well as of the quantum kernel method.

What problem does this paper attempt to address?

The paper is primarily dedicated to addressing the effective quantum embedding of classical data in Quantum Machine Learning (QML), particularly for supervised learning tasks. Specifically, the research focuses on the following points: 1. **Importance of Quantum Embedding**: The paper emphasizes the critical role of quantum embedding in quantum machine learning, especially when dealing with classical data. The quality of quantum embedding directly affects the model's expressiveness, generalization ability, and training capability. 2. **Limitations of Existing Methods**: Traditional quantum embedding methods often fail to ensure that the embedded quantum states have high data separability, which is crucial for improving classification performance. Additionally, using parameterized quantum circuits for quantum embedding increases the depth and number of gates in the quantum circuit, which is infeasible on Noisy Intermediate-Scale Quantum (NISQ) devices and can easily lead to the "barren plateau" phenomenon during training. 3. **Proposed Method - Neural Quantum Embedding (NQE)**: To address the above issues, the authors propose Neural Quantum Embedding (NQE), a method that combines classical neural networks with quantum embedding techniques. The goal of NQE is to optimize quantum embedding to maximize the trace distance between quantum states of different classes, thereby improving data distinguishability and lowering the empirical risk bound. This method avoids the problems present in traditional methods, such as increased quantum circuit depth and training obstacles. 4. **Theoretical Foundation and Experimental Validation**: The paper theoretically analyzes how NQE improves the empirical risk bound in quantum binary classification tasks by enhancing the trace distance between quantum states. Through experiments on IBM quantum devices, the authors demonstrate significant improvements in multiple metrics, including training accuracy, generalization ability, training stability, and robustness to noise. 5. **Comparison with Other Methods**: NQE is also compared with existing trainable quantum embedding methods. The results show that NQE effectively improves data separability and performs excellently in classification accuracy in both noise-free and noisy environments. In summary, this paper aims to overcome the limitations of existing quantum embedding methods by introducing Neural Quantum Embedding technology, thereby enhancing the performance of quantum machine learning algorithms in processing classical data.

Neural Quantum Embedding: Pushing the Limits of Quantum Supervised Learning

Quantum Algorithms for Feedforward Neural Networks

Quantum Neural Network for Quantum Neural Computing

Quantum Data Encoding: A Comparative Analysis of Classical-to-Quantum Mapping Techniques and Their Impact on Machine Learning Accuracy

Quantum Hamiltonian Embedding of Images for Data Reuploading Classifiers

Quantum Visual Feature Encoding Revisited

Exploring Quantum-Enhanced Machine Learning for Computer Vision: Applications and Insights on Noisy Intermediate-Scale Quantum Devices

Quantum enhancements for deep reinforcement learning in large spaces

Neural Networks for Programming Quantum Annealers

Quantum Deep Learning Neural Networks

Accurate Image Multi-Class Classification Neural Network Model with Quantum Entanglement Approach

Harnessing Quantum Extreme Learning Machines for image classification

Quantum Neural Network Classifiers: A Tutorial

Quantum Pointwise Convolution: A Flexible and Scalable Approach for Neural Network Enhancement

Leveraging Data Locality in Quantum Convolutional Classifiers

Leveraging Pre-Trained Neural Networks to Enhance Machine Learning with Variational Quantum Circuits

Integrated Encoding and Quantization to Enhance Quanvolutional Neural Networks

Problem-Dependent Power of Quantum Neural Networks on Multi-Class Classification

Problem-Dependent Power of Quantum Neural Networks on Multiclass Classification

Quantum-enhanced neural networks in the neural tangent kernel framework