Model Components
Discriminator (in GAN)
The discriminator is a classifier trained to output a probability that its input is real (from the training data) rather than fake (from the generator). As training progresses, the discriminator becomes more sophisticated, providing better training signal for the generator.
In the ideal GAN equilibrium (Nash equilibrium), the generator produces data indistinguishable from real, and the discriminator outputs 0.5 for all inputs. Practical training rarely reaches this point; techniques like gradient penalties and spectral normalization improve stability.
Authority Links
Related Terms
Model Components
Generative Adversarial Network (GAN)
Framework training two competing networks—a generator and discriminator—to produce realistic synthetic data.
Model Components
Generator
GAN component that creates synthetic data instances designed to be indistinguishable from real data.
Techniques & Methods
Adversarial Training
Training AI models on challenging, adversarially crafted inputs to improve robustness and reliability.
Model Components
Generative Model
AI model that learns to generate new data instances resembling the training distribution.