An Analysis of Recurrent Neural Network variants to predict yield of Wheat crop in Punjab region

Abstract

Yield prediction of a crop prior to harvest is an important aspect of agriculture. The alarming rate at which the population of the world is increasing is making it necessary to take measures to increase the yield of major food crops. The advancements in the area of data sciences especially in the field of machine learning and deep learning has come up as a major breakthrough in the area of crop yield prediction. The ability of deep learning to perform accurate predictions even in the areas with huge and diverse natured data has made it the most sort after technique for predictions in agriculture. There are many agriculture based regions in India where the efficiency of these advanced techniques is still not fully utilized in the field of crop yield prediction. The objective of the present study is to evaluate the efficiency of variants of Recurrent Neural Network, to predict the yield of Wheat crop for a region of Punjab state in India.