Crop yield forecasting using neural networks

Abstract

The crop production forecasting has become an important issue, now, as it is a key factor for our economy and sustainable development on account of increased demand of the food grains with growing population. It helps farmers and government to develop a better post-harvest management at local / regional / national level, e.g., transportation, storage, distribution. Additionally, it helps farmers to plan next year's crop and government to plan import/export strategies. This work is based on the yield forecasting of the pearl millet (bajra) in the Jaipur region of Rajasthan, India. The proposed method uses a back propagation artificial neural network to forecast current yield of the crop with respect to the environmental factors using time series data. The obtained results are encouraging and much better in comparison to a recent fuzzy time series based methods for forecasting. © 2013 Springer International Publishing.