摘要
本文根据中止决策的实质及模式识别的基本原理 ,探讨了中止 R& D项目的一般模式识别程序。同时 ,基于小波基具有很强的自适应数据和函数变化的能力 ,进一步研究给出了一种基于小波网络的 R& D项目中止决策分析模式识别新方法 ,并在最小均方能量准则下 ,采用相应的共轭梯度学习算法求解子波函数线性组合的尺度和时延参数 ,以及神经网络的权值。本研究不仅克服了现有 R& D项目中止决策分析技术在构造与求解模型方面的某些局限性 ,并且为 R& D项目的跟踪管理提供了一种新的理论依据 ,为具体决策单位和政府有关部门对正在实施的 R&
According to the significance and charaters of R& D project termination decision and basic principle of pattern identification,the paper discussed general pattern identification proceduces about R& D project termination decision and proposed a new mode identification method of R& D project temination decision based on the property that the Wavelet can smooth datum and adapt to the changes of function.Furthmore,the dialation and shifts which are the parameters of wavelet function and the weight of neural network structure are calculated by conjugate gradient algorithm in term s of the latest m ean squared energy criterion.The study not only overcome the difficulties and drawbacks in the exsting method of R& D project term ination decision,but provide also a new theoretical basis for following management of R& D projects and a set of operational decision method and procedure for specific decision makers or government departments.
出处
《预测》
CSSCI
2001年第1期35-39,14,共6页
Forecasting
基金
国家自然科学基金资助项目! (79770 10 5)
重庆市重点软科学资助项目! (556 9)
关键词
R&D项目
中止决策
小波网络模型识别
共轭梯度法
项目管理
R& D project
termination decision
Wavelet network pattern identification
conjugate gradient algorithm.
