A comparison of feed-forward and recurrent neural networks in time series forecasting

Brezak, Danko and Baček, Tomislav and Majetić, Dubravko and Kasać, Josip and Novaković, Branko (2012) A comparison of feed-forward and recurrent neural networks in time series forecasting. = A comparison of feed-forward and recurrent neural networks in time series forecasting. In: 2012 IEEE Conference on Computational Intelligence for Financial Engineering and Economics, CIFEr 2012, 29-30.03.2012., New York City, NY, United States.

575899.A_Comparison_of_Feed-forward_and_Recurrent_Neural_Networks_in_Time_Series_Forecasting.pdf - Submitted Version Jezik dokumenta:English

Download (437kB) | Preview
Official URL: https://doi.org/10.1109/CIFEr.2012.6327793


Forecasting performances of feed-forward and recurrent neural networks (NN) trained with different learning algorithms are analyzed and compared using the Mackey-Glass nonlinear chaotic time series. This system is a known benchmark test whose elements are hard to predict. Multi-layer Perceptron NN was chosen as a feed-forward neural network because it is still the most commonly used network in financial forecasting models. It is compared with the modified version of the so-called Dynamic Multi-layer Perceptron NN characterized with a dynamic neuron model, i.e., Auto Regressive Moving Average filter built into the hidden layer neurons. Thus, every hidden layer neuron has the ability to process previous values of its own activity together with new input signals. The obtained results indicate satisfactory forecasting characteristics of both networks. However, recurrent NN was more accurate in practically all tests using less number of hidden layer neurons than the feed-forward NN. This study once again confirmed a great effectiveness and potential of dynamic neural networks in modeling and predicting highly nonlinear processes. Their application in the design of financial forecasting models is therefore most recommended.

Item Type: Conference or Workshop Item (Lecture)
Keywords (Croatian): autoregressive moving average; benchmark tests; chaotic time series; dynamic neural networks; dynamic neurons; feed-forward; financial forecasting; forecasting performance; hidden layer neurons; multi layer perceptron; nonlinear process; recurrent NN; time series forecasting; artificial intelligence; benchmarking; learning algorithms; neurons; recurrent neural networks; time series; forecasting
Divisions: 900 Department of Robotics and Production System Automation > 910 Chair of Engineering Automation
Indexed in Web of Science: Yes
Indexed in Current Contents: No
Citations JCR: 0 (19.09.2018.)
Date Deposited: 30 Apr 2015 09:00
Last Modified: 29 Oct 2018 13:52
URI: http://repozitorij.fsb.hr/id/eprint/3906

Actions (login required)

View Item View Item


Downloads per month over past year