Premium
Neural Networks for Forecasting: An Introduction
Author(s) -
Kohzadi Nowrouz,
Boyd Milton S.,
Kaastra Iebeling,
Kermanshahi Bahman S.,
Scuse David
Publication year - 1995
Publication title -
canadian journal of agricultural economics/revue canadienne d'agroeconomie
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.505
H-Index - 37
eISSN - 1744-7976
pISSN - 0008-3976
DOI - 10.1111/j.1744-7976.1995.tb00135.x
Subject(s) - artificial neural network , computer science , scope (computer science) , artificial intelligence , process (computing) , autoregressive integrated moving average , machine learning , demand forecasting , time series , operations research , engineering , programming language , operating system
Neural networks are nonlinear mapping structures based on the study of the human brain. They have been shown to be universal and highly flexible junction approximators to any data‐generating process. Therefore, they are powerful models for forecasting purposes, especially when the underlying data‐generating processes are unknown. However, the appropriate design of the network's architecture and learning rules are crucial for obtaining satisfactory results. This study discusses the scope and limitations of neural networks for forecasting problems and provides an example by designing a neural network for forecasting. It is argued that statistical theory can offer some suggestions for designing an optimal network architecture. An example comparing a neural network and ARIM model for forecasting weekly corn prices 1974 through 1993 is provided. Results show the neural network model to be more accurate than the ARIMA.