ATime Series Forecasting Using Neural Network by Modified Adaptive Learning Rates and Initial Values


The Transactions of the Korea Information Processing Society (1994 ~ 2000), Vol. 5, No. 10, pp. 2609-2614, Oct. 1998
10.3745/KIPSTE.1998.5.10.2609,   PDF Download:

Abstract

In this work, we consider the forecasting performance between neural network and Box-Jenkins method for time series data. A modified learning process is developed for neural approach at time series data, ie. properly adaptive learning rates selecting by orthogonal arrays and dynamic selecting of initial values using Easton's controller box. We can obtain good starting points with dynamic graphics approach. We use real data sets for this study : the Wolf yearly sunspot numbers between 1700 and 1988.


Statistics
Show / Hide Statistics

Statistics (Cumulative Counts from September 1st, 2017)
Multiple requests among the same browser session are counted as one view.
If you mouse over a chart, the values of data points will be shown.


Cite this article
[IEEE Style]
Y. Y. Chang and L. S. Duck, "ATime Series Forecasting Using Neural Network by Modified Adaptive Learning Rates and Initial Values," The Transactions of the Korea Information Processing Society (1994 ~ 2000), vol. 5, no. 10, pp. 2609-2614, 1998. DOI: 10.3745/KIPSTE.1998.5.10.2609.

[ACM Style]
Yoon Yeo Chang and Lee Sung Duck. 1998. ATime Series Forecasting Using Neural Network by Modified Adaptive Learning Rates and Initial Values. The Transactions of the Korea Information Processing Society (1994 ~ 2000), 5, 10, (1998), 2609-2614. DOI: 10.3745/KIPSTE.1998.5.10.2609.