Friday, 3 October 2025

Data Driven Approach for Modeling and Forecasting of Maize Yield in India | Chapter 14 | Agricultural Sciences: Techniques and Innovations Vol. 4

 

The demand for food grain crops is increasing at a rapid rate worldwide. To meet the global food demand, efforts are needed for the enhancement of crop yield through improved varieties, policy support, subsidies, resource allocation, market development, and farmers’ motivation towards the cultivation of profitable crops. However, the yield of agricultural crops is influenced by several extraneous factors like climate change, pest attacks, resource scarcity, and land acquisition for construction and urbanisation. This study deals with modelling and forecasting of maize yield in India. The study encompasses temporal data from 1954 to 2023, allowing for a data-driven analysis of trends in maize yield over the years in India. The methodology involved the fitting of several conventional autoregressive integrated moving average (ARIMA) models. The accuracy of the fitted models was evaluated using model fit statistics criteria viz., akaike information criterion (AIC), root mean square error (RMSE) and mean absolute percentage error (MAPE). The best-fitted model was ARIMA (2,1,0) with drift, having an AIC value of 894.95, an RMSE value of 148.05, and an MAPE value of 7.71%. Furthermore, a comparative assessment of the conventional fitted models was made with the automated model, viz., ARIMA (1,1,2) with drift, which was o.btain.d on using function in R-studio. The diagnostic checking of residuals of the generated models was made using the Ljung-Box test. It was revealed that the Ljung-Box test statistic (Q) achieved a p-value greater than 0.05 for residuals of each fitted model, which indicated the acceptance of the null hypothesis (H0), i.e., the residuals of the various generated models were uncorrelated. The analytical results revealed that ARIMA (1,1,2) with a drift model slightly outperformed ARIMA (2,1,0) with a drift. The forecast values of maize yield for five successive years (viz., 2024-2028) were obtained with 80% and 95% prediction intervals utilising ARIMA (1,1,2) with drift model. The outcomes of the analysis reported a significantly rising trend of maize yield over the recent years, which indicated a favourable sign for policymakers and scientists regarding the formulation of strategies related to agricultural trade and nutritional security.

 

 

Author(s) Details

Manish Kumar
Department of Agricultural Statistics, Acharya Narendra Deva University of Agriculture & Technology, Ayodhya, India.

 

Shiv Kumar Rana
Department of Agricultural Statistics, Acharya Narendra Deva University of Agriculture & Technology, Ayodhya, India.

 

Please see the book here :- https://doi.org/10.9734/bpi/asti/v4/6379

No comments:

Post a Comment