计算机工程与应用 ›› 2011, Vol. 47 ›› Issue (32): 224-227.

• 工程与应用 • 上一篇    下一篇

离散过程神经网络在时间序列预测中的应用

刘丽杰1,李盼池2,李 欣2,张 强2   

  1. 1.黑龙江八一农垦大学 信息技术学院,黑龙江 大庆 163319
    2.东北石油大学 计算机与信息技术学院 黑龙江 大庆 163318
  • 收稿日期:1900-01-01 修回日期:1900-01-01 出版日期:2011-11-11 发布日期:2011-11-11

Application of discrete process neural network in time series prediction

LIU Lijie1,LI Panchi2,LI Xin2,ZHANG Qiang2   

  1. 1.School of Information Technology,HLJ August First Land Reclamation University,Daqing,Heilongjiang 163319,China
    2.School of Computer and Information Technology,Northeast Petroleum University,Daqing,Heilongjiang 163318,China
  • Received:1900-01-01 Revised:1900-01-01 Online:2011-11-11 Published:2011-11-11

摘要: 为解决复杂时间序列的预测问题,针对目前过程神经网络的输入为多个连续的时变函数,而许多实际问题的输入为多个序列的离散值,提出一种基于离散输入的过程神经网络模型及学习算法;并以太阳黑子数实际数据为例对太阳黑子数时间序列进行预测,仿真结果表明该模型具有很好的逼近和预测能力。

关键词: 过程神经元网络, 学习算法, 时间序列预测, 太阳黑子数

Abstract: By now,the input of Process Neural Network(PNN) is multiple continuous time-varying function and the input for practical problems is discrete value of multiple series,a PNN model and learning algorithm is presented based on discrete input to solve the problem of complex time series prediction.The algorithm takes sunspot number as example to predict sunspot number time series,and the simulation results show that the model produces good ability for approximation and prediction.

Key words: process neural networks, learning algorithm, time series predication, sunspot number