2024, Vol. 9, Special Issue 1

The population of the world is increasing at a rapid rate and more specifically in India resulting in increase of the food per capita requirement. Wheat plays an important role in ensuring global food with the demand of over 40 percent for the year 2030. In order to ascertain this requirement a reliable forecasting is essential for decision-makers to plan adequate policies and to establish the necessary logistical resources. In this sense, the Artificial Neural Networks was used to predict wheat production, productivity and cultivated area of wheat in India and compare the same with the classical methods of time series data. The time series data of 60 years from 1961 to 2021 of India was collected from the Food and Agriculture Organization of United Nations. The data was analysed using classical methods namely linear, exponential, logarithm, polynomial and power function. During the study the respective trendlines and equations were drawn using MS Excel for classical methods and also evaluation matrices parameter namely coefficient of determination (R2), Mean Squared Error (MSE) and Mean Absolute Error (MAE) were estimated to determine the suitability of the model with respect to the time series data. The same data was then used in prediction analysis using Artificial Neural Networks (ANN) on R studio software and similar evaluation metrices were estimated. From the study it has been observed that the ANN proved to be the best model in predicting the production, productivity and cultivated area of the wheat with the highest R values of 0.9 and least error values of both MSE and MAE. For validation of data with the actual, the activation functions like linear and tangent function predicted the values nearer to the actual values. Furthermore, artificial neural networks (ANNs) are considered a dependable model for forecasting time series data, enabling stakeholders to plan in advance for the rising demands.