摘要
提出了一种线性时间序列与神经网络非线性时间序列相结合的"两级模式识别"的分级预测与分类的技术:首先采用线性时间序列分析方法对Web集群的服务模式进行总体预测与分类,在总体模式预测与分类的基础上采用神经网络非线性算法进行QoS模式精确识别。理论与实验分析结果证实了该方法的有效性。
The article introduces a level predicting and classifying method named 'two level pattern recognization', combining the linear time series and the nonlinear time series of neural network. We can predict and classify overally the service pattern of Web clusters by linear time series, then recognize exactly the QoS pattern by nonlinear algorithm of neural network. The method has been proved effective by theoretic analysis and experiments.
出处
《计算机应用》
CSCD
北大核心
2003年第9期5-7,共3页
journal of Computer Applications
基金
湖南省自然科学基金项目 (0 2JJY2 0 97)
关键词
QoS模式
时间序列
神经网络
预测与分类
QoS pattern
time series
neural network
predict and classification
