185
Kevin
Personal Homepage
Paper Publications
一种量子条件生成对抗网络算法
Hits :

Journal:电子学报

Key Words:量子生成对抗网络;条件信息;W态编码;参数化量子电路

Abstract:量子生成对抗网络是量子机器学习算法领域研究热点之一,但其生成过程具有较大的随机性,不太适用于现实场景. 为了解决该问题,提出了一种生成过程可控的量子条件生成对抗网络(Quantum Conditional Generative Adversarial Network,QCGAN)算法,其中条件信息采用one-hot形式进行多粒子W态编码,并通过向生成器和判别器输
入条件信息达到稳定模型生成过程的目的. 性能评估表明,与经典GAN、CGAN相比,本算法可生成离散数据,且将时间复杂度从 O(N2 )降为 O(N);与带条件约束的量子生成对抗网络 QuGAN 相比,QCGAN 消耗更少的量子资源 . 最后,以BAS(3,3)数据集和量子混合态生成为例,选用PennyLane平台进行仿真实验,结果表明QCGAN算法经过训练可有效收敛到Nash均衡点,进而验证了算法的实验可行性.

Indexed by:Journal paper

Document Code:0372

Document Type:J

Volume:50

Issue:7

Page Number:1586-1593

Translation or Not:no

Date of Publication:2022-07-13

Personal information

Professor
Supervisor of Doctorate Candidates
Supervisor of Master's Candidates

Gender : Male

Education Level : With Certificate of Graduation for Doctorate Study

Degree : Doctoral Degree in Engineering

Status : 在岗

School/Department : 软件学院

Discipline:Other Specialties in Software Engineering
Computer Science and Technology

Business Address : 信息科技大楼(临江楼)A1107-1108

Contact Information : 18795809602

PostalAddress : 科技信息大楼(临江楼)A1107-1108

Telephone : 18795809602

Email : wenjiel@163.com

You are visitors

The Last Update Time : ..


Copyright©2019 Nanjing University of Information Science and Technology·Network Information Center

MOBILE Version