Application of referential ensemble learning techniques to predict the density of rainfall

We propose a novel neural-network based model for predicting rainfall in any given city in India. Our initial implementation primarily focuses on a rural city in Bihar, a state which experiences erratic rainfall behavior. We start with simpler models based on time-series analysis and linear regression, ultimately combining them using a novel look back-window method. The initial results are promising and comparable with the state-of-the-art methods published in this domain. It is conjectured that atmospheric pressure and winds are highly correlated to rainfall, and we are trying to obtain data for the same to use in our model. In addition, we wish to extend the work to predict droughts as well. Observing and analyzing the changes in the crop yield based on the precipitation is also something we intend to do. India is so powerless against environmental change is on the grounds that it is a vast nation with numerous living in destitution, deficient base, and is subjected to unpredictable rainstorm, flooding, rising ocean levels, and higher temperatures. The rainstorm season is essential to the Indian economy in light of the fact that numerous Indians are agrarian. Thusly India's rainstorm will radically influence the destiny of the agrarian segment and the general population dependant on it. Environmental change is going to keep on creating whimsical extremes all through the storm season. Readiness for climate inconsistencies achieved by environmental change is along these lines key to ensure the lives of the Indian individuals and the development of the Indian economy.