English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools annual

Global: Zendy Free Plan

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Generative adversarial networks (GAN) present state-of-the-art results in thegeneration of samples following the distribution of the input dataset. However,GANs are difficult to train, and several aspects of the model should bepreviously designed by hand. Neuroevolution is a well-known technique used toprovide the automatic design of network architectures which was recentlyexpanded to deep neural networks. COEGAN is a model that uses neuroevolutionand coevolution in the GAN training algorithm to provide a more stable trainingmethod and the automatic design of neural network architectures. COEGAN makesuse of the adversarial aspect of the GAN components to implement coevolutionarystrategies in the training algorithm. Our proposal was evaluated in theFashion-MNIST and MNIST dataset. We compare our results with a baseline basedon DCGAN and also with results from a random search algorithm. We show that ourmethod is able to discover efficient architectures in the Fashion-MNIST andMNIST datasets. The results also suggest that COEGAN can be used as a trainingalgorithm for GANs to avoid common issues, such as the mode collapse problem.

COEGAN: Evaluating the Coevolution Effect in Generative Adversarial Networks