2016 GANSforSequencesofDiscreteEleme

• (Kusner & Hernndez-Lobato, 2016) ⇒ Matt J. Kusner, and Jose Miguel Hernndez-Lobato. (2016). “GANS for Sequences of Discrete Elements with the Gumbel-softmax Distribution". In: arXiv:1611.04051.

Subject Headings: GANS for Sequences of Discrete Elements with the Gumbel-softmax Distribution.

Notes

• Other Version(s) and Link(s):
  • DBLP: https://dblp.org/rec/html/journals/corr/KusnerH16

Cited By

• Google Scholar: ~ 145 Citations.
• Semantic Scholar: ~ 147 Citations.
• MS Academic: ~ 137 Citations.

Quotes

Abstract

Generative Adversarial Networks (GAN) have limitations when the goal is to generate sequences of discrete elements. The reason for this is that samples from a distribution on discrete objects such as the multinomial are not differentiable with respect to the distribution parameters. This problem can be avoided by using the Gumbel-softmax distribution, which is a continuous approximation to a multinomial distribution parameterized in terms of the softmax function. In this work, we evaluate the performance of GANs based on recurrent neural networks with Gumbel-softmax output distributions in the task of generating sequences of discrete elements.

BibTeX

@article{2016_GANSforSequencesofDiscreteEleme,
  author    = {Matt J. Kusner and
               Jose  Miguel Hernndez-Lobato},
  title     = {GANS  for Sequences of Discrete Elements with the Gumbel-softmax
               Distribution},
  journal   = {CoRR},
  volume    = {abs/1611.04051},
  year      = {2016},
  url       = {http://arxiv.org/abs/1611.04051},
  archivePrefix = {arXiv},
  eprint    = {1611.04051},
}

---

	Author	volume	Date Value	title	type	journal	titleUrl	doi	note	year
2016 GANSforSequencesofDiscreteEleme	Matt J. Kusner Jose Miguel Hernndez-Lobato			GANS for Sequences of Discrete Elements with the Gumbel-softmax Distribution						2016