Improved Crop Yield Forecasting using Wavelet-Based Denoising and Artificial Neural Networks Framework
166 / 115
Keywords:
Hybrid model; Neural network; Pulse yield;Wavelets and volatility.Abstract
Agricultural crop yields are often non-stationary, noisy and influenced by multiple environmental factors, which limits the effectiveness of traditional
statistical models. Advanced hybrid approaches that can capture both linear and non-linear patterns, along with multi-scale characteristics, are required
for reliable forecasting. This study aimed to develop and evaluate a hybrid wavelet-based forecasting framework by integrating the Maximal Overlap
Discrete Wavelet Transform (MODWT) with Artificial Neural Network (ANN) and Autoregressive Integrated Moving Average (ARIMA) models
for predicting crop yield series of Total Pulses, Gram, and Tur in India. Annual yield data from 1950 to 2023 were analysed using a hybrid modelling
framework. The MODWT was applied to decompose the original time series into multi-resolution components, followed by denoising through
thresholding techniques. The decomposed components were then modelled using ANN and ARIMA. Optimal wavelet filters and decomposition
levels were obtained based on predictive performance. The results showed that hybrid wavelet-based models significantly outperformed standalone
ARIMA and ANN models. This study contributes to the literature by systematically optimizing wavelet filters and decomposition levels within a
hybrid modelling framework and demonstrates the effectiveness of combining wavelet-based denoising with machine learning and statistical models
for non-stationary agricultural time series data.
Downloads
References
1. Alhnaity, B., Kollias, S., Leontidis, G., Jiang, S., Schamp, B. and
2. Pearson, S. (2021). An autoencoder wavelet based deep neural
3. network with attention mechanism for multi-step prediction of
4. plant growth. Information Sciences, 560, 35-50
5. Anjoy, P., Paul, R.K., Sinha, K., Paul, A.K. and Ray, M. (2017). A hybrid
6. wavelet based neural networks model for predicting monthly WPI
7. of pulses in India. Indian Journal of Agricultural Sciences, 87(6),
8. 834-839.
9. Bakulin, D.V., Beletskaya, N.V., Gorshkov, N.N., Djioeva, M.I.,
10. Mokhnatkina, U.S., Petrusevich, D.A. and Polyakov, A.A. (2025).
11. Hybrid Models of Time Series Using Wavelet Transform. Journal
12. of Communications Technology and Electronics, 70(7), 309-318
13. Burrus S., Gopinath R., and Guo H. (1998). Introduction to Wavelets
14. and Wavelet Transforms: A Primer. Recherche, 67.
15. Das, P., Jha, G.K., Lama, A., and Parsad, R. (2023). Crop yield
16. prediction using hybrid machine learning approach: a case study
17. of lentil (Lens culinaris Medik.). Agriculture, 13(3), 596.
18. Daubechies, I. (1992) Ten lectures on wavelets. Society for industrial
19. and applied mathematics.
20. Hounmenou, C.G., Gneyou, K.E. and Kakai, R.G. (2021). A Formalism
21. of the General Mathematical Expression of Multilayer Perceptron
22. Neural Networks.
23. Huang, W., Huang, Z., Matei, M. and Wang, T. (2012). Price volatility
24. forecast for agricultural commodity futures: the role of high
25. frequency data. Romanian Journal of Economic Forecasting.
26. Khosravi, S., Nikoo, M.R., Zamani, M.G., Barzegar, R., Najafipour, P.
27. and Gandomi, A.H., (2026). Enhancing long-term water quality
28. forecasting with a hybrid deep-learning approach integrating
29. MODWT, CNN, and GRU. Hydrological Sciences Journal, 71(2),
30. 270-293.
31. Lindfield, G. and Penny, J. (2019). Analysing data using discrete
32. transforms. Numerical Methods, 383-431
33. Lionel, B.M., Musabe, R., Gatera, O. and Twizere, C., (2025). A
34. comparative study of machine learning models in predicting crop
35. yield. Discover Agriculture, 3(1),151.
36. Mila, F.A. and Parvin, M.T., (2019). Forecasting area, production and
37. yield of onion in Bangladesh by using ARIMA model. Asian
38. Journal of Agricultural Extension, Economics & Sociology,
39. 37(4),1-12.
40. Mitra, D. and Paul, R.K. (2017). Hybrid time-series models for
41. forecasting agricultural commodity prices. Model Assisted
42. Statistics and Applications, 12,255–264
43. Musunuru, N., Yu, M. and Larson, A. (2013). Forecasting volatility
44. of returns for corn using GARCH Models. Texas Journal of
45. Agriculture and Natural Resources, 26:42-55.
46. 148
47. Md Yeasin et al. / Journal of the Indian Society of Agricultural Statistics 135–148
48. Naik, R.T. (2025). Comparative Study of Wavelet–ANN and Wavelet
49. ARIMA Models for Groundwater Level Forecasting. International
50. Journal on Science and Technology, 16(4).
51. Oikonomidis, A., Catal, C. and Kassahun, A. (2022). Hybrid deep
52. learning-based models for crop yield prediction. Applied artificial
53. intelligence, 36(1), 2031822.
54. Paidipati, K. and Banik, A. (2020). Forecasting of rice cultivation
55. in India–a comparative analysis with ARIMA and LSTM-NN
56. models. EAI Endorsed Transactions on Scalable Information
57. Systems, 7(24).
58. Paul, R. K., Prajneshu, and Ghosh, H. (2009). GARCH Nonlinear Time
59. Series Analysis for Modeling and Forecasting of India’s Volatile
60. Spices Export Data. Journal of the Indian Society of Agricultural
61. Statistics, 63(2), 123-131.
62. Paul, R.K., Ghosh, H. and Prajneshu (2014). Development of out
63. of-sample forecast formulae for ARIMAX-GARCH model and
64. their application. Journal of the Indian Society of Agricultural
65. Statistics, 68(1), 85-92.
66. Paul, R.K., Paul, A.K., and Bhar, L.M. (2019). Wavelet-based
67. combination approach for modelling sub-divisional rainfall in
68. India. Theoretical and Applied Climatology, 139(3-4), 949-964.
69. Paul, R.K. and Garai, S. (2021). Performance comparison of wavelets
70. based machine learning technique for forecasting agricultural
71. commodity prices. Soft Computing, 25(20), 12857-12873
72. Paul, R.K. and Garai, S. (2022). Wavelets based artificial neural network
73. technique for forecasting agricultural prices. Journal of the Indian
74. Society for Probability and Statistics, 23(1), 47-61.
75. Paul, R.K., Garai, S., and Yeasin, M. (2025). Wavelet-SVR: A novel
76. wavelet-based support vector regression algorithm for time series
77. forecasting. International Journal of Agricultural Sustainability,
78. 23(1), 1-18.
79. Paul, R.K., Yeasin, M. and Manjunatha, B. (2025). WaveDeepCrop:
80. wavelet enhanced deep learning framework for noise-resilient
81. crop yield forecasting. Proceedings of the Indian National Science
82. Academy, 91(4), 1310-1323.
83. Rakshit, D. and Paul, R.K. (2024). Hybrid ARFIMA-LRNN Model for
84. Forecasting Commodity Prices. Journal of the Indian Society of
85. Agricultural Statistics, 78(3), 221–229.
86. Rani M. (2025). Application of wavelet transform in agricultural
87. sciences. International Journal of Statistics and Applied
88. Mathematics,10(8), 33-36.
89. Shao, Z., Chao, F., Yang, S.L. and Zhou, K.L. (2017). A review of
90. the decomposition methodology for extracting and identifying
91. the fluctuation characteristics in electricity demand forecasting.
92. Renewable and Sustainable Energy Reviews, 75, 123-136.
93. Sharif, I. and Khare, S. (2014). Comparative analysis of Haar and
94. Daubechies wavelet for hyper spectral image classification. The
95. International Archives of the Photogrammetry, Remote Sensing
96. and Spatial Information Sciences, 40, 937-941.
97. Shawon, S.M., Ema, F.B., Mahi, A.K., Niha, F.L. and Zubair, H.T.
98. (2025). Crop yield prediction using machine learning: An
99. extensive and systematic literature review. Smart Agricultural
100. Technology, 10, 100718