Rice Crop Yield Prediction Using Random Forest and Deep Neural Network - An Integrated Approach

Abstract

Prediction of crop yields with the help of modern techniques like machine learning algorithms is always in demand to increase the productivity of crops under various environmental conditions. This prediction can help the farmers and other stakeholders to make better decisions on yielding to crops. That is, when to cultivate and what to cultivate. Machine learning algorithms have been used for various crop yield predictions such as rice, maize, wheat, etc. Among the crops, rice is always in demand where countries like India. Especially, Tamilnadu which is one of the states in India is giving more contribution to rice productivity compare with other states. Though the cultivation of rice is increasing in this state, various environmental conditions affect the yield of rice productivity. Thus, in this paper, an integrated approach using random forest and the deep neural network has been proposed. The ultimate aim of this approach is to improve the rice crop prediction under various environmental conditions. To achieve this, this approach uses an ensemble technique by combining the significance of random forest algorithm as well as deep neural network algorithm. The accuracy of this approach has been evaluated by various statistical indicators such as recall, accuracy, precision, and F1 score. The results witnessed that the proposed approach gives better prediction accuracy compares with traditional random forest and deep neural network algorithms.