Adversarially Regularized Autoencoders for Generating Discrete Structures
arXiv: Learning, Volume abs/1706.04223, 2017.
Generative adversarial networks are an effective approach for learning rich latent representations of continuous data, but have proven difficult to apply directly to discrete structured data, such as text sequences or discretized images. Ideally we could encode discrete structures in a continuous code space to avoid this problem, but it i...More
Full Text (Upload PDF)
PPT (Upload PPT)