REAL TIME BIG DATA ANALYTICS FOR AGRICULTURAL LAND HOTSPOT PREDICTION

Agricultural land hotspots are those that experienced sudden fall in vegetation due to climatic changes that occurred over years. Rather than taking measures after the land has been identified as a hotspot, it is better taking preventive measures to avoid the land from becoming a hotspot. Once the regions (hotspot) data is gathered, the model is trained to predict the probability of any land to become a hotspot. The prediction is carried out by analyzing the big data of hotspot regions. The machine learns from the observations as learning algorithm identifies patterns in the data set. Satellite land image classification is done by classifying land area to forest, water and land areas and making decisions based on the land cover.