Data analytics on agrometeorological parameters for building a utility system for farmer community

Abstract

The day-to-day evolution of the world is throwing many challenges, thereby demanding the humans to be on par with the modern tools and technologies that are on offer. These technologies are contributing to the overall growth across all the domains. One such domain that is highly affected by the modern tools and techniques is Agriculture. Gone are the days, when only a handful methodologies and tools were utilized to understand about agriculture. Although the farmers are sceptical to the modernization, the end result has been encouraging at many a times. With the use of analytics, one can get to know more about the type of soil, crops and fertilizers, amount of water to be utilized depending on the climatic conditions and thereby have an effective yield. In this work, the elementary task of portraying the effect of climatic conditions on the production of different varieties of crops is carried out. Also, the application of Multivariate Linear Regression and Artificial Neural Network (ANN) techniques provides a significance outcome in the prediction of yield with the focus on Ragi and Rice. The work provides an accuracy of 65% with the 3-fold cross validation technique and 68% accuracy with ANN model.