Analytics Based on Artificial Neural Network: A Case Study Based on Iowa Corn Yield Forecasting
Abstract
This work is an application of computing techniques to solve real world problems. Agriculture has been playing a crucial role in the growth of world economy and has been associated with production of basic food demands. The World�s food demand is increasing day by day with the population but the cultivable land which is a limited resource is not increasing in proportion with the food demand. In order to maximize the production and improvise the quality of crop, various firms have come up with variety of fertilizes, genetic seeds, modern equipment, etc. While the technological advancement is still going on, there is another area of interest where most of the technological firms are now focusing, which is predicting the approximate crop yield in the different geographical locations.In order to address the above problem, we proposed a model to understand the effects of various crop yield influencing parameters such as field location, soil properties, availability water in the ground, ground slope and climate conditions such as rainfall and temperature. We proposed this ANN model to predict corn yield of IOWA region with these influencing parameters which will help the various agricultural firms to give the recommendations to their growers to maximize the crop yield. According to our knowledge, this is the first implementation of ANN model to predict corn yield of IOWA region with these set of data features. In this work, the ANN model produced more consistent yield prediction and it resulted in R2 of 0.70 and RMSE of around 1750.
Collections
- OSU Theses [15752]