改进的Hamming神经网络在R&D项目中止决策中的应用  

Application of Improved Hamming Neural Networks to R&D Project Termination Decisions

作　　者：官建成[1] 刘权[1] 

机构地区：[1]北京航空航天大学管理学院,北京100083

出　　处：《控制与决策》2001年第6期882-885,889,共5页Control and Decision

基　　金：国家自然科学基金项目 (79770 0 0 9);航空预研项目

摘　　要：针对 Ham ming网络的匹配网络中输出节点的阈值 ,采用一种既能反映模式向量又能反映输入向量的调整方式 ,使改进的 Hamm ing网络能够测度基于欧氏距离的向量间的贴近度。探讨了改进的Hamming网络在 R& D项目中止决策领域应用的可行性 ,同时对改进的 Hamm ing神经网络算法迭代的收敛性给出了证明。案例研究表明 ,改进后的 Ham ming神经网络能有效地对 R&Aiming at the threshold of the output nodes in the first layer of the Hamming neural networks, an adjusting method is used so that the output can reflect not only the state of the model vectors but also that of those input vectors. The improved Hamming neural networks can measure the degree of the closeness based on the Euclid distance between the vectors. The feasibility of the application of the improved Hamming nerual networks to the R&D project termination decision is discussed. A proof for the convergence of the iterative arithmetic is given. Case study shows that the improved Hamming neural networks can produce valid and interesting results.

关 键 词：中止决策 HAMMING神经网络 R&D项目 技术创新 企业 

分 类 号：F273.1[经济管理—企业管理] TP183[经济管理—国民经济]