.png){kind=icon}
返回
New parameterized quantum gates design and efficient gradient solving based on variational quantum classification algorithm
.png){kind=icon}
OA
AI
DOI：10.1088/2632-2153/ada0a4.png)
.png){kind=icon}
.png){kind=icon}
分享.png){kind=icon}
收藏
摘要
En 中文
Currently, variational quantum classification algorithms (VQCAs) generally rely on traditional optimization techniques such as Powell and SLSQP in the parameter optimization session. However, the performance of these methods shows limitations in practical applications. Although the parameter-shift rule can efficiently compute the parameter gradient with quantum circuits, it needs to run the quantum circuit twice repeatedly, which significantly reduces the computation efficiency. In order to overcome this challenge, this paper innovatively integrates the principle of unitary operation in quantum mechanics with the technical characteristics of superconducting quantum chips and elaborately designs some new parameterized quantum gates (PQGs). These PQGs strictly follow the rules of unitary operation, which ensures the stability and accuracy of quantum state evolution while realizing an efficient solution to the parameter gradient. Especially for the gradient calculation of a single qubit and single-angle PQGs, the new method can be completed with only a single quantum circuit run, which greatly improves the computation efficiency. Experimental validation on benchmark datasets such as breast cancer and iris shows that the method proposed in this paper exhibits excellent performance on quantum classification tasks. Compared with the parameter-shift rule, the computation efficiency of the new method is improved by 40%. And the classification accuracy, precision, and other key performance metrics are improved by an average of 5% in comparison with traditional optimization algorithms. This work not only enriches the methodology of quantum machine learning theoretically but also demonstrates its remarkable superiority in practical applications, which indicates that the method has great potential in scientific research and industrial applications.
Keyword:
VQCAs
parameters optimization
PQGs
quantum-classical hybrid algorithms
期刊
M
IF:
4.6
论文数:
1.1K
被引数:
3.9K
机构
暂无机构信息
引用论文
A variational eigenvalue solver on a photonic quantum processor光子量子处理器上的变分特征值求解器.png)
NATURE COMMUNICATIONS
OAAI
NATURE COMMUNICATIONSIF15.7
OAAI分享收藏Review series on helminths, immune modulation and the hygiene hypothesis: Mechanisms underlying helminth modulation of dendritic cell functionImmunologyOAAI
ImmunologyIF0
OAAI分享收藏分享收藏Predicting retrosynthetic pathways using transformer-based models and a hyper-graph exploration strategyCHEMICAL SCIENCEOAAI
CHEMICAL SCIENCEIF7.4
OAAI分享收藏
查看更多内容