TL;DR: qGANs combine quantum circuits with the GAN framework, using a quantum generator (parameterized circuit) and either quantum or classical discriminator. They can learn and generate complex probability
qGANs combine quantum circuits with the GAN framework, using a quantum generator (parameterized circuit) and either quantum or classical discriminator. They can learn and generate complex probability distributions, with potential advantages in modeling high-dimensional quantum states. IBM has demonstrated qGANs for financial data loading and distribution learning.
Type
Machine Learning / Generative
Complexity
Heuristic; potential advantage in high-dimensional distributions
Application
Financial modeling, data augmentation, distribution learning, anomaly detection