A Machine Learning Perspective for Predicting Agricultural Droughts

Abstract

Drought affects a large number of people and cause more loss to society compared to other natural disasters. Tamilnadu is a drought disaster-prone state in India. Data mining in agriculture is a very recent research topic. This technique is used for agriculture in data mining. A related, but not equivalent term is precision agriculture. The frequent occurrence of drought posses an increasingly severe threat to the Tamilnadu agricultural production. Drought also has very complex phenomenon that is difficult to accurately quantify because it’s immense spatial and temporal variability. In the Existing system, ISDI model construction was implemented for evaluating the accuracy and the effectiveness. This model application using a variety of methods and data, there is still some work to be done in our future research because of the complex spatial and temporal characteristics of drought. To overcome imitation, it assesses performance of measuring the drought by using the Spatial and temporal characteristics of information’s. The dataset is collected from different regions and also collects the time varying information’s from the dataset. In this, the drought conditions was predicted by using the supervised learning mechanism. It can be implemented by using the Bayesian supervised machine learning algorithm. Through this project accuracy and performance can be achieved, and also performance and effectiveness can be improved.