backpropagation
Ciência da Computação
Verbete 1 de 1
substantivo
Contexto: "Monte Carlo Tree Search Algorithm (Selection, Expansion, Simulation, Backpropagation), was employed in 2006. Monte Carlo Search Tree, unlike FSM, has more sophisticated and varied rooted derivatives. Each state has its own likelihood percentage, which will be determined based on a subset of current conditions. NPCs will seem more alive if this method is used since they will have several levels of considerations to make decisions and actions."
Fonte: Yakan, SA. Analysis of Development of Artificial Intelligence in the Game Industry. Disponível em: Analysis of Development of Artificial Intelligence in the Game Industry | International Journal of Cyber and IT Service Management (iaic-publisher.org). Acesso em: 09 jun. 2023.
Fonte: Yakan, SA. Analysis of Development of Artificial Intelligence in the Game Industry. Disponível em: Analysis of Development of Artificial Intelligence in the Game Industry | International Journal of Cyber and IT Service Management (iaic-publisher.org). Acesso em: 09 jun. 2023.
Termo equivalente: retropropagação; retropropagação de erro.
Definição: "Backpropagation is a common method of practicing a neural net in which the basic system output is compared to the desired output, and the system is fine-tuned until the difference between the two is diminished."
Fonte: Rathore, C.S.; Unal, Aynur; Rajasekera, Jay. Early detection of Heart Diseases using ANN (Time and Cost Efficient). Disponível em: Early detection of Heart Diseases using ANN (Time and Cost Efficient) (researchgate.net). Acesso em: 08 jun. 2023.
Fonte: Rathore, C.S.; Unal, Aynur; Rajasekera, Jay. Early detection of Heart Diseases using ANN (Time and Cost Efficient). Disponível em: Early detection of Heart Diseases using ANN (Time and Cost Efficient) (researchgate.net). Acesso em: 08 jun. 2023.
Definição em português: "Retropropagação é um processo pelo qual os pesos são corrigidos pelo método do gradiente decrescente de forma a que o erro seja mínimo. Erro é a diferença entre o desempenho desejado e o obtido na(s) saída(s) da rede neural artificial (RNA)."