Quantum Generative Adversarial Networks (QGANS)

EasyChair Preprint 14860

Abstract

Quantum Generative Adversarial Networks (QGANs) represent a significant advancement in the intersection of quantum computing and machine learning. By leveraging the principles of quantum mechanics, QGANs aim to overcome the limitations of classical Generative Adversarial Networks (GANs) in terms of computational efficiency and model expressiveness. In a QGAN framework, the generator and discriminator are implemented using quantum circuits, which allows for the encoding of complex probability distributions and the generation of high-dimensional data with potentially superior fidelity. This paper explores the theoretical foundations of QGANs, highlighting how quantum superposition and entanglement can enhance the learning capabilities of GANs. We also review recent developments in quantum algorithms that facilitate the training of QGANs and discuss the challenges associated with their implementation on current quantum hardware. By comparing QGANs with classical GANs, we identify key areas where quantum-enhanced methods offer significant advantages and outline future research directions to address the practical and theoretical obstacles in deploying QGANs at scale.

Keyphrases: Generative Adversarial Networks, natural language text, quantum

@booklet{EasyChair:14860,
  author    = {Favour Olaoye and Kaledio Potter},
  title     = {Quantum Generative Adversarial Networks (QGANS)},
  howpublished = {EasyChair Preprint 14860},
  year      = {EasyChair, 2024}}