基于神经网络的参数-费用模型研究

讨论了采用用神经网络理论来建立参数－费用模型的方法，给出了几种地空导弹武器系统的特征参数与研制费用，对神经网络法与线性回归法所得的结果进行了比较。比较结果表明，神经网络法比线性回归法精确。

STUDY ON PAREMETER-COST MODEL BASED ON NEURAL NETWORK

This paper discusses the methods to establish a paremeter-cost model by usingneural network theory , gives caracteristic paremeters and development costs of several types of air-defense missile weapon systems, and compares the results obtained from neural network with that from linear regression .The comparing results show that neural network method is moreaccurate than linear regression method .