skip to main content
Idioma:
Primo Search
Clear Search Box
Search in: Busca Geral
Busca Avançada
Busca por Índices

PHYLOGENETIC REPLAY LEARNING IN DEEP NEURAL NETWORKS

Glafkides, Jean ; Sher, Gene ; Akdag, Herman

Jordanian Journal of Computers and Information Technology, 2022-09, Vol.8 (3), p.1-231 [Periódico revisado por pares]

Amman: Scientific Research Support Fund of Jordan Princess Sumaya University for Technology

Texto completo disponível

Citações Citado por
  • Título:
    PHYLOGENETIC REPLAY LEARNING IN DEEP NEURAL NETWORKS
  • Autor: Glafkides, Jean ; Sher, Gene ; Akdag, Herman
  • Assuntos: Automation ; Back propagation ; Deep learning ; Genetic algorithms ; Information technology ; Methods ; Mutation ; Network management systems ; Neural networks ; neuroevolution ; Optimization algorithms ; phylogenetic replay learning ; Phylogenetics ; vanishing gradient
  • É parte de: Jordanian Journal of Computers and Information Technology, 2022-09, Vol.8 (3), p.1-231
  • Descrição: Though substantial advancements have been made in training deep neural networks, one problem remains, the vanishing gradient. The very strength of deep neural networks, their depth, is also unfortunately their problem, due to the difficulty of thoroughly training the deeper layers due to the vanishing gradient. This paper proposes "Phylogenetic Replay Learning", a learning methodology that substantially alleviates the vanishing gradient problem. Unlike the residual learning methods, it does not restrict the structure of the model. Instead, it leverages elements from Neuroevolution, transfer learning, and layer by layer training. We demonstrate that this new approach is able to produce a better performing model, and by calculating Shannon Entropy of weights, we show that the deeper layers are trained much more thoroughly and contain statistically significantly more information than when a model is trained in a traditional brute force manner...
  • Editor: Amman: Scientific Research Support Fund of Jordan Princess Sumaya University for Technology
  • Idioma: Inglês;Árabe
Voltar para lista de resultados

  • Realização: ABCD-USP USP   Logos de Redes Sociais: Freepik

Buscando em bases de dados remotas. Favor aguardar.