Constructing an appropriate Neural Network for maximising Sugarcane Yield in a particular region

Abstract

With vast degree of industrialization there is a severe amount of depletion in the Levels of soil fertility which in turn could impact the growth of crops. Care needs to be taken to identify the right soil conducive for growing various kinds of crops. This work focuses on the ideal variety of sugar cane crop that could be grown in a particular type of soil in India which in turn could maximize the overall yield of sugarcane. An additional parameter dealing with the amount of sugar content has also been taken into consideration. The choice of an appropriate decision would go a long way in reducing avoidable losses in terms of both capital and effort. With this view in perspective, a wide range of artificial neural networks have been constructed and the results have been compiled. Besides, choices of appropriate training function and learning rates have been made.