Comparative performance of wavelet-based neural network approaches
- 241 Downloads
An agriculture-dominated developing country like India has been always in need of efficient and reliable time series forecasting methodologies to describe various agricultural phenomenons, whereas agricultural price forecasting continue to be the challenging areas in this domain. The observed features of many temporal price data set constitute complex nonlinearity, and modeling these features often go beyond the capability of Box–Jenkins autoregressive integrated moving average methodology. Moreover, despite the popularity and sheer power of traditional neural network model, the empirical forecasting performance of this model has not been found satisfactory in all cases. To address the problem, wavelet-based modeling approach is recently upsurging. Present study discusses two wavelet-based neural network approaches envisaging monthly wholesale onion price of three markets, namely Bangalore, Hubli, and Solapur. Wavelet-based decomposition makes it possible to describe the useful pattern of the series from both global as well as local aspects and found to be highly proficient in denoising and capturing the inherent pattern of the series through a distinctive approach. Besides, wavelet method can also be used as a tool for function approximation. The improvement upon time-delay neural network also be made up to a great extent through using wavelet-based approaches as exhibited through proper empirical evidence.
KeywordsARIMA MODWT Nonlinearity TDNN Wavelet
We would like to express our sincere thanks and gratitude to the anonymous reviewers for their valuable suggestions that helped us a lot in improving this manuscript.
Compliance with ethical standards
Conflict of interest
The authors declare that there is no conflict of interests regarding the publication of this paper.
- 2.Anjoy P, Paul RK, Sinha K, Paul AK, Ray M (2017) A hybrid wavelet based neural networks model for predicting monthly wpi of pulses in India. Indian J Agric Sci 87(6):834–839Google Scholar
- 8.Granger CWJ, Anderson AP (1978) Introduction to bilinear time series models. Vandenhoeck and Ruprecht, GottingenGoogle Scholar
- 14.Mohammadi K, Eslami HR, Dardashti D (2005) Comparison of regression, ARIMA and ANN models for reservoir inflow forecasting using snowmelt equivalent (a case study of Karaj). J Agric Sci Technol 7:17–30Google Scholar
- 15.Pacelli V, Bevilacqua V, Azzollini M (2011) An artificial neural network model to forecast exchange rates. J Intell Learn Syst Appl 3:57–69Google Scholar
- 16.Paul RK, Prajneshu GH (2013) Statistical modelling for forecasting of wheat yield based on weather variables. Indian J Agric Sci 83(2):180–183Google Scholar
- 17.Paul RK, Das MK (2013) Forecasting of average annual fish landing in Ganga Basin. Fish Chimes 33(3):51–54Google Scholar
- 19.Paul RK, Alam W, Paul AK (2014) Prospects of livestock and dairy production in India under time series framework. Indian J Anim Sci 84(4):130–134Google Scholar
- 21.Paul RK, Gurung B, Paul AK (2015) Modelling and forecasting of retail price of arhar dal in Karnal, Haryana. Indian J Agric Sci 85(1):69–72Google Scholar
- 22.Paul RK, Sinha K (2016) Forecasting crop yield: a comparative assessment of ARIMAX and NARX model. RASHI 1(1):77–85Google Scholar