【作者】 叶德谦； 马剑；

【Author】 Ye Deqian~1 Ma Jian~2 (ICDZ,Yanshan University,Qinhuangdao 066004)

【机构】 燕山大学中德信息技术研究所；

【摘要】 本文针对传统专家系统和神经网络在预测方面的不足,提出了一种基于模糊理论-BP 神经网络-专家系统混合时序预测的模型。通过对外部定性参数的模糊处理,有效地解决了外部知识的量化。采用神经网络和专家系统的结合方法,不仅克服了专家知识获取的瓶颈问题,而且达到了微观数据预测和宏观趋势预测目的。该模型在实际应用中取得了较好效果。更多还原

【Abstract】 This paper proposes the hybrid time series prediction model based on fuzzy theory-BP neural network-expert system aiming at the fault of traditional expert system and neural network in prediction field. Quantification of exterior information is effectively solved by fuzzy processing of exterior qualitative- parameter. The combination of neural network and expert system is used,it not only overcomes the"bottle-neck"problems in acquiring knowledge,but also attains the purpose of micro data-prediction and macro trend-prediction.The model is tested and the expected results are obtained.更多还原