摘要
讨论了神经元算法对线性可分模式空间的适应性,将其应用于某合金钢生产优化区的寻找,并将之与主成分分析法所得结果作了比较,研究结果表明,神经元算法简单、可靠性高,是一种自学习的模式识別的好方法。
The suitability of neuron algorithm to linear seperatable pattern space was discussed. It was applied to find the optimal condition in production of some alloyed steels. A comparison was made between neuron algorithm and traditional pattern recognition technique. The results show that the neuron algorithm is a simple, reliable algorithm and self adaptive pattern recognition technique.
出处
《金属学报》
SCIE
EI
CAS
CSCD
北大核心
1992年第11期B487-B489,共3页
Acta Metallurgica Sinica
关键词
神经元
模式识别
合金钢
neuron, pattern recognition, principal compment analysis, alloy steel
