.png){kind=icon}
.png){kind=icon}
Return
Optimizer-dependent generalization bound for quantum neural networks
.png){kind=icon}
PRE
AI
DOI：10.1007/s42484-025-00318-9.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Quantum neural networks (QNNs) play a pivotal role in addressing complex tasks within quantum machine learning, analogous to classical neural networks in deep learning. Understanding the generalization properties of QNNs is essential for ensuring reliable performance and informing their design in quantum machine learning applications. In this paper, we investigate the generalization properties of QNNs through the lens of learning algorithm stability, circumventing the need to explore the entire hypothesis space and providing insights into how classical optimizers influence QNN performance. By establishing a connection between QNNs and quantum combs, we examine the general behaviors of QNN models from a quantum information theory perspective. Leveraging the uniform stability of the stochastic gradient descent algorithm, we propose a generalization error bound determined by the number of trainable parameters, data uploading times, dataset dimension, and classical optimizer hyperparameters. Numerical experiments validate this comprehensive understanding of QNNs and align with our theoretical conclusions. As the first exploration into understanding the generalization capability of QNNs from a unified perspective of design and training, our work offers practical insights for applying QNNs in quantum machine learning.
Keywords:
Quantum machine learning
Quantum neural networks
Generalization
Stability
Journal
Q
IF:
4.4
Papers:
429
Citations:
796