Deep Adversarial Belief Networks
This provides a new foundation for training DBNs with enhanced scalability and flexibility, potentially benefiting researchers in machine learning and applied fields, though it appears incremental as it adapts existing GAN concepts.
The authors tackled the problem of training deep belief networks (DBNs) without back-propagation by developing an adversarial framework that replaces the generator in GANs with a DBN, resulting in a scalable and generalizable method applicable to various domains like computer vision and neuroscience.
We present a novel adversarial framework for training deep belief networks (DBNs), which includes replacing the generator network in the methodology of generative adversarial networks (GANs) with a DBN and developing a highly parallelizable numerical algorithm for training the resulting architecture in a stochastic manner. Unlike the existing techniques, this framework can be applied to the most general form of DBNs with no requirement for back propagation. As such, it lays a new foundation for developing DBNs on a par with GANs with various regularization units, such as pooling and normalization. Foregoing back-propagation, our framework also exhibits superior scalability as compared to other DBN and GAN learning techniques. We present a number of numerical experiments in computer vision as well as neurosciences to illustrate the main advantages of our approach.