Machine learning based Pedantic Analysis of Predictive Algorithms in Crop Yield Management
M. Chandraprabha, Rajesh Kumar Dhanaraj
Abstract
Predictive analytics is a statistical technique used to forecast and investigate the development from past chronological data or to extract the information from data. With the help of rising technologies like predictive analytics in data mining, machine learning combining with Internet of Things [IoT], the major challenges in crop yield can be solved and pave way to earn profit. Machine learning means the process of making the system to learn from the previous experiences that help in prediction. In this paper, an conjectural evaluation on diverse prediction algorithms like support vector machines (SVM), recurrent neural networks (RNN), K nearest neighbour regression (KNN-R), Naive Bayes, BayesNet, support vector regression (SVR) etc., is done and its performance are described on the basis of error rates and accuracy level in crop yield. BayesNet shows the higher accuracy of about 97.53% and RNN has less percentage error rates that dominate other algorithms in harvest prediction.