摘要
本文讨论了人工神经网络(ANN)解决问题的能力,从广泛的角度讨论了容量问题,推广与学习问题,深入研究了ANN通过学习解决问题的计算复杂性,以及解决实际问题时困难所在。
The problem solving ability of ANN is discussed in this paper, problem of capacity,generalization and learning is studied from wide view of point. the computational complexity of ANN learning is discussed in deep. Difficulties in solving practical problems by ANN is also pointed out.
出处
《电子学报》
EI
CAS
CSCD
北大核心
1995年第5期63-67,共5页
Acta Electronica Sinica
基金
国家攀登计划
认知科学(神经网络)重大关键项目
关键词
神经网络
人工智能
Neural network
Computational complexity
Learning