Embedding Learning in Hybrid Quantum-Classical Neural Networks
arXiv:2204.04550 · doi:10.1109/QCE53715.2022.00026
Abstract
Quantum embedding learning is an important step in the application of quantum machine learning to classical data. In this paper we propose a quantum few-shot embedding learning paradigm, which learns embeddings useful for training downstream quantum machine learning tasks. Crucially, we identify the circuit bypass problem in hybrid neural networks, where learned classical parameters do not utilize the Hilbert space efficiently. We observe that the few-shot learned embeddings generalize to unseen classes and suffer less from the circuit bypass problem compared with other approaches.
arXiv comment8 pages, 11 figures
References in corpus (17)
- Prototypical Networks for Few-shot Learning
- Quantum algorithm for solving linear systems of equations
- Supervised learning with quantum enhanced feature spaces
- Quantum support vector machine for big data classification
- Quantum Convolutional Neural Networks
- The power of quantum neural networks
- Power of data in quantum machine learning
- Quantum autoencoders for efficient compression of quantum data
- The Expressive Power of Parameterized Quantum Circuits
- Hybrid Quantum-Classical Convolutional Neural Networks
- Supervised quantum machine learning models are kernel methods
- Quantum Variational Autoencoder
- Simulation of low-depth quantum circuits as complex undirected graphical models
- Representation Learning via Quantum Neural Tangent Kernels
- An adaptive algorithm for quantum circuit simulation
- Tensor Network Quantum Simulator With Step-Dependent Parallelization
- Performance Evaluation and Acceleration of the QTensor Quantum Circuit Simulator on GPUs