Título: Treinamento de Redes Neurais para Tarefas de Inspeção Mediante Reforços de Múltiplos Críticos
Autores: Agra, Paulo Baz; Yoneyama, Takashi; Nascimento Júnior, Cairo Lúcio
Resumo: The main objective of this work is to investigate the concepts of game theory in the context of reinforcement learning with multiple critics. The neural net is assumed to perform an inspection task, where desirable input patterns should produce high ‘accept’ output values while undesirable input patters are expected to yield high ‘reject’ output values. The multiple critics may or may not cooperate, so that game theoretic situations arise. In this context, the stochastic learning automata is shown to converge to Nash equilibrium points or Pareto solutions, depending on the nature of the information state.
Palavras-chave:
Páginas: 5
Artigo em pdf: 4cbrn_085.pdf
Arquivo BibTex: 4cbrn_085.bib