摘要
基于ART-2神经网络分类器算法是在自适应共振理论的基础上改进而来的,该方法能够对造船中间产品的特性分析样本进行分析判断,实现输入数据的自动分类与识别。提出了中间产品成组特性分析的模型,并对神经网络结构和改进的算法作了详细介绍。实验证明分类的效果比较准确,并可对同族产品的相似程度加以控制,使分组更加合理。
The classifier of ART-2 artificial neural network is proposed on the basis of the Adaptive Resonance Theory. It can automatically classify and identify the input data by analyzing the characteristics of interim products in shipbuilding. In this paper, the model of characteristic analysis of interim product families is provided, and ART-2 network structure and modified algorithm are in detail introduced respectively. An example based on the algorithm proves that the classified results are correct and the similar degree of the cognate product can be controlled. In this way, the grouping is reasonable.
出处
《中国造船》
EI
CSCD
北大核心
2006年第2期108-113,共6页
Shipbuilding of China
