WHICH METHODOLOGY IS BETTER FOR COMBINING LINEAR AND NONLINEAR MODELS FOR TIME SERIES FORECASTING

Both theoretical and empirical findings have suggested that combining different models can be an effective way to improve the predictive performance of each individual model. It is especially occurred when the models in the ensemble are quite different. Hybrid techniques that decompose a time series into its linear and nonlinear components are one of the most important kinds of the hybrid models for time series forecasting. Several researches in the literature have been shown that these models can outperform single models. In this paper, the predictive capabilities of three different models in which the autoregressive integrated moving average (ARIMA) as linear model is combined to the multilayer perceptron (MLP) as nonlinear model, are compared together for time series forecasting. These models are including the Zhang’s hybrid ANNs/ARIMA, artificial neural network (p,d,q), and generalized hybrid ANNs/ARIMA models. The empirical results with three well-known real data sets indicate that all of these methodologies can be effective ways to improve forecasting accuracy achieved by either of components used separately. However, the generalized hybrid ANNs/ARIMA model is more accurate and performs significantly better than other aforementioned models.

[1]  C. Granger,et al.  Efficient Market Hypothesis and Forecasting , 2002 .

[2]  K. Hipel,et al.  Time series modelling of water resources and environmental systems , 1994 .

[3]  Bruce Mizrach,et al.  Multivariate nearest‐neighbour forecasts of ems exchange rates , 1992 .

[4]  R. Shibata Selection of the order of an autoregressive model by Akaike's information criterion , 1976 .

[5]  J. Schmidt,et al.  Comparison of the forecasting accuracy of neural networks with other established techniques , 1991, Proceedings of the First International Forum on Applications of Neural Networks to Power Systems.

[6]  Kin Keung Lai,et al.  A novel nonlinear ensemble forecasting model incorporating GLAR and ANN for foreign exchange rates , 2005, Comput. Oper. Res..

[7]  Minvydas Ragulskis,et al.  Non-uniform attractor embedding for time series forecasting by fuzzy inference systems , 2009, Neurocomputing.

[8]  T. Taskaya-Temizel,et al.  Are ARIMA neural network hybrids better than single models? , 2005, Proceedings. 2005 IEEE International Joint Conference on Neural Networks, 2005..

[9]  Manoochehr Ghiassi,et al.  A dynamic architecture for artificial neural networks , 2005, Neurocomputing.

[10]  Chang-Hua Hu,et al.  An effective hybrid approach based on grey and ARMA for forecasting gyro drift , 2008 .

[11]  Gwilym M. Jenkins,et al.  Time series analysis, forecasting and control , 1972 .

[12]  H. Brian Hwarng,et al.  Insights into neural-network forecasting of time series corresponding to ARMA(p,q) structures , 2001 .

[13]  R. D. Jones,et al.  A Neural Net Model for Prediction , 1994 .

[14]  Kenneth S. Rogoff,et al.  Exchange rate models of the seventies. Do they fit out of sample , 1983 .

[15]  R. Clemen Combining forecasts: A review and annotated bibliography , 1989 .

[16]  Michael Y. Hu,et al.  Forecasting with artificial neural networks: The state of the art , 1997 .

[17]  Krzysztof J. Cios,et al.  Time series forecasting by combining RBF networks, certainty factors, and the Box-Jenkins model , 1996, Neurocomputing.

[18]  Gwilym M. Jenkins,et al.  Time series analysis, forecasting and control , 1971 .

[19]  W. Li,et al.  On a mixture autoregressive model , 2000 .

[20]  Tommaso Minerva,et al.  Building ARMA Models with Genetic Algorithms , 2001, EvoWorkshops.

[21]  Tugba Taskaya-Temizel,et al.  2005 Special Issue: A comparative study of autoregressive neural network hybrids , 2005 .

[22]  T. Rao,et al.  An Introduction to Bispectral Analysis and Bilinear Time Series Models , 1984 .

[23]  Mehdi Khashei,et al.  A new hybrid artificial neural networks and fuzzy regression model for time series forecasting , 2008, Fuzzy Sets Syst..

[24]  P. A. Fishwick Neural network models in simulation: a comparison with traditional modeling approaches , 1989, WSC '89.

[25]  Spyros Makridakis,et al.  Why combining works , 1989 .

[26]  Lennart Ljung,et al.  System Identification: Theory for the User , 1987 .

[27]  Paul A. Fishwick,et al.  Neural Network Models In Simulation: A Comparison With Traditional Modeling Approaches , 1989, 1989 Winter Simulation Conference Proceedings.

[28]  Yongqiang Tang,et al.  A consistent nonparametric Bayesian procedure for estimating autoregressive conditional densities , 2007, Comput. Stat. Data Anal..

[29]  Guoqiang Peter Zhang,et al.  An empirical investigation of bias and variance in time series forecasting: modeling considerations and error evaluation , 2003, IEEE Trans. Neural Networks.

[30]  Guoqiang Peter Zhang,et al.  Time series forecasting using a hybrid ARIMA and neural network model , 2003, Neurocomputing.

[31]  A. Lapedes,et al.  Nonlinear Signal Processing Using Neural Networks , 1987 .

[32]  Daihai He,et al.  Chaotic oscillations and cycles in multi-trophic ecological systems. , 2007, Journal of theoretical biology.

[33]  Ashu Jain,et al.  Hybrid neural network models for hydrologic time series forecasting , 2007, Appl. Soft Comput..

[34]  Michele Marchesi,et al.  A hybrid genetic-neural architecture for stock indexes forecasting , 2005, Inf. Sci..

[35]  Mehdi Khashei,et al.  An artificial neural network (p, d, q) model for timeseries forecasting , 2010, Expert Syst. Appl..

[36]  王正華,et al.  A hybrid SARIMA and support vector machines in forecasting the production values of the machinery industry in Taiwan , 2007 .

[37]  Mehdi Khashei,et al.  A novel hybridization of artificial neural networks and ARIMA models for time series forecasting , 2011, Appl. Soft Comput..

[38]  Jens Ove Riis,et al.  A hybrid econometric—neural network modeling approach for sales forecasting , 1996 .

[39]  T. Ouarda,et al.  Generalized autoregressive conditional heteroscedasticity modelling of hydrologic time series , 2012 .

[40]  Fang-Mei Tseng,et al.  Combining neural network model with seasonal time series ARIMA model , 2002 .

[41]  Mascha C. van der Voort,et al.  Combining kohonen maps with arima time series models to forecast traffic flow , 1996 .

[42]  J. Padmore,et al.  A threshold model for the French franc/deutschmark exchange rate , 1996 .

[43]  Chee Peng Lim,et al.  Predicting drug dissolution profiles with an ensemble of boosted neural networks: a time series approach , 2003, IEEE Trans. Neural Networks.

[44]  Jih-Jeng Huang,et al.  Model identification of ARIMA family using genetic algorithms , 2005, Appl. Math. Comput..

[45]  Kuan-Yu Chen,et al.  A hybrid SARIMA and support vector machines in forecasting the production values of the machinery industry in Taiwan , 2007, Expert Syst. Appl..

[46]  C. Granger,et al.  An introduction to bilinear time series models , 1979 .

[47]  Kishan G. Mehrotra,et al.  Forecasting the behavior of multivariate time series using neural networks , 1992, Neural Networks.

[48]  R. Engle Autoregressive conditional heteroscedasticity with estimates of the variance of United Kingdom inflation , 1982 .

[49]  Francesco Giordano,et al.  Forecasting nonlinear time series with neural network sieve bootstrap , 2007, Comput. Stat. Data Anal..

[50]  Wang Yao-wu,et al.  Neural Network Forecasting of the Production Level of Chinese Construction Industry , 2003 .

[51]  J. Keith Ord,et al.  Automatic neural network modeling for univariate time series , 2000 .

[52]  G. Yule Why do we Sometimes get Nonsense-Correlations between Time-Series?--A Study in Sampling and the Nature of Time-Series , 1926 .

[53]  Guoping Xia,et al.  An investigation and comparison of artificial neural network and time series models for Chinese food grain price forecasting , 2007, Neurocomputing.

[54]  D. J. Reid Combining Three Estimates of Gross Domestic Product , 1968 .

[55]  Ping-Feng Pai,et al.  A hybrid ARIMA and support vector machines model in stock price forecasting , 2005 .

[56]  Miki Haseyama,et al.  An ARMA order selection method with fuzzy reasoning , 2001, Signal Process..

[57]  Elmar Steurer,et al.  Much ado about nothing? Exchange rate forecasting: Neural networks vs. linear models using monthly and weekly data , 1996, Neurocomputing.

[58]  Claas de Groot,et al.  Analysis of univariate time series with connectionist nets: A case study of two classical examples , 1991, Neurocomputing.

[59]  David Horn,et al.  Combined Neural Networks for Time Series Analysis , 1993, NIPS.

[60]  Lyle H. Ungar,et al.  Neural network forecasting of short, noisy time series , 1992 .

[61]  Thomas Kolarik,et al.  Time series forecasting using neural networks , 1994, APL '94.

[62]  Guoqiang Peter Zhang,et al.  A neural network ensemble method with jittered training data for time series forecasting , 2007, Inf. Sci..

[63]  Leandro dos Santos Coelho,et al.  Computational intelligence approaches and linear models in case studies of forecasting exchange rates , 2007, Expert Syst. Appl..

[64]  Jiwen Dong,et al.  Time-series forecasting using flexible neural tree model , 2005, Inf. Sci..

[65]  Andreas S. Weigend,et al.  Time Series Prediction: Forecasting the Future and Understanding the Past , 1994 .

[66]  Chakradhara Panda,et al.  Forecasting exchange rate better with artificial neural network , 2007 .

[67]  Marie Cottrell,et al.  Neural modeling for time series: A statistical stepwise method for weight elimination , 1995, IEEE Trans. Neural Networks.

[68]  Ray Tsaih,et al.  Forecasting S&P 500 stock index futures with a hybrid AI system , 1998, Decis. Support Syst..

[69]  David Hsieh Chaos and Nonlinear Dynamics: Application to Financial Markets , 1991 .

[70]  Paul A. Fishwick,et al.  Feedforward Neural Nets as Models for Time Series Forecasting , 1993, INFORMS J. Comput..

[71]  James W. Denton,et al.  How Good Are Neural Networks for Causal Forecasting , 1995 .

[72]  Paul A. Fishwick,et al.  Time series forecasting using neural networks vs. Box- Jenkins methodology , 1991, Simul..

[73]  J. M. Bates,et al.  The Combination of Forecasts , 1969 .

[74]  M. Kendall,et al.  A Study in the Analysis of Stationary Time-Series. , 1955 .

[75]  Kyung-shik Shin,et al.  A hybrid approach based on neural networks and genetic algorithms for detecting temporal patterns in stock markets , 2007, Appl. Soft Comput..

[76]  Clifford M. Hurvich,et al.  Regression and time series model selection in small samples , 1989 .

[77]  H. Tong,et al.  Threshold Autoregression, Limit Cycles and Cyclical Data , 1980 .

[78]  P.-A. Cornillon,et al.  Forecasting time series using principal component analysis with respect to instrumental variables , 2008, Comput. Stat. Data Anal..

[79]  Mehdi Khashei,et al.  Improvement of Auto-Regressive Integrated Moving Average models using Fuzzy logic and Artificial Neural Networks (ANNs) , 2009, Neurocomputing.