Sensitivity Analysis Applied to Artificial Neural Networks for Forecasting Time Series
Abstract
This paper presents a novel procedure known as sensitivity analysis applied to a multilayer perceptron (MLP), which allows the most relevant lagged terms in time series forecasting to be identified. Second, this paper conducts a comparison of forecasting accuracy between the neural network model resulting from applying the sensitivity analysis to the network model derived from the traditional procedure and the classic ARIMA model – using the time series corresponding to the number of passengers in transit through the Balearic Islands. Our findings demonstrate that a neural network derived from sensitivity analysis provides the greatest forecasting accuracy.
References
(2005). Stock market forecasting: Artificial neural network and linear regression comparison in an emerging market. Journal of Financial Management and Analysis, 18, 18–33.
(2007). Comparison of ARIMA, neural networks, and hybrid models in time series: Tourist arrival forecasting. Journal of Statistical Computation & Simulation, 77, 29–53.
(2007). Forecasting airborne pollen concentration time series with neural and neuro-fuzzy models. Expert Systems with Applications, 32, 1218–1225.
(1999). Input feature selection for real-time transient stability assessment for artificial neural network (ANN) using ANN sensitivity analysis. In , Proceedings of the 21st International Conference on Power Industry Computer Applications (pp. 295–300). Piscataway, NJ: IEEE.
(1992). First- and second-order methods for learning: Between steepest descent and Newtons method. Neural Computation, 4, 141–166.
(1989). What size net gives valid generalization? Neural Computation, 1, 151–160.
(1995). Neural networks for pattern recognition. Oxford, UK: Oxford University Press.
(1976). Time series analysis: Forecasting and control. San Francisco: Holden Day.
(1982). Evaluation of extrapolative forecasting methods: Results of academicians and practitioners. Journal of Forecasting, 1, 215–217.
(1994). Sensitivity analysis on neural networks for meteorological variable forecasting. In , Proceedings of IEEE Workshop on Neural Networks for Signal Processing (pp. 587–595). New York: IEEE.
(1989). Approximation by superpositions of a sigmoidal function. Mathematical Control, Signal and Systems, 2, 303–314.
(1989). On the approximate realization of continuous mappings by neural networks. Neural Networks, 2, 183–192.
(1999). Time series prediction with genetic-algorithm designed neural networks: An empirical comparison with modern statistical models. Computational Intelligence, 15, 171–184.
(2001). Investigation of the nonlinearity in the heart rate dynamics. Medical Engineering and Physics, 23, 111–115.
(1989). Multilayer feedforward networks are universal approximators. Neural Networks, 2, 359–366.
(2007). Analysis of bioclimatic time series and their neural network-based classification to characterize drought risk patterns in South Italy. International Journal of Biometeorology, 51, 253–263.
(1996). Designing a neural network for forecasting financial and economic time series. Neurocomputing, 10, 215–236.
(1999). A neural network model to forecast Japanese demand for travel to Hong Kong. Tourism Management, 20, 89–97.
(1993). Practical neural networks recipes in C++. London: Academic Press.
(1995). Advanced algorithms for neural networks: A C++ sourcebook. New York: Wiley.
(2003). Numeric sensitivity analysis applied to feedforward neural networks. Neural Computing and Applications, 12, 119–125.
(2006). Designing an artificial neural network for forecasting tourism time series. Tourism Management, 27, 781–790.
(2006). Modeling and forecasting the energy consumption in Taiwan using artificial neural networks. The Journal of American Academy of Business, 8, 113–119.
(1996). Using a neural network to forecast visitor behavior. Annals of Tourism Research, 23, 151–164.
(1999). Critical input data channels selection for progressive work exercise test by neural network sensitivity analysis. In , IEEE International Conference on Acoustics, Speech, and Signal Processing (pp. 1097–1100). Piscataway, NJ: IEEE.
(1986). Learning internal representations by error propagation. In , Parallel distributed processing (pp. 318–362). Cambridge, MA: MIT Press.
(1998). Neural network data analysis using Simulnet. New York: Springer-Verlag.
(1978). Estimating the dimension of a model. The Annals of Statistics, 6, 461–464.
(2005). Forecasting gaming referenda. Annals of Tourism Research, 32, 127–149.
. (1998). Neural Connection 2.1. Chicago: SPSS Inc.
(1991). Input layer optimization of neural networks by sensitivity analysis and its application to recognition of numerals. Transactions of the Institute of Electrical Engineers Japan, 111, 36–44.
(1989). Neural computing: Theory and practice. New York: Van Nostrand Reinhold.
(1997). A comparative analysis of neural networks and statistical methods for predicting consumer choice. Marketing Science, 16, 370–391.
(2000). Neural network training for prediction of climatological time series, regularized by minimization of the generalized cross-validation function. Monthly Weather Review, 128, 1456–1473.
(2007). An investigation and comparison of artificial neural network and time series models for Chinese food grain price. Neurocomputing, 70, 2913–2923.
