A Comprehensive Review on AI Based Model for Rainfall Prediction

Bhavini K. Shah, Dr. Kamlendu S. Pandey

Abstract

Our nation’s economy is heavily dependent on agriculture and industries. To generate a profit, we should
rely substantially on the availability of water. But the outcome is severely hampered by the irregularity of
rainfall and the depletion of available water supplies. In 2014, India earned net $8 billion from $304 billion
trade in commercial services. On the other hand, agricultural trade of $56 billion fetched as much as $18
billion in trade surplus. This is because while in services trade imports account for a lion’s share, in
agriculture imports component is negligible since basic resources such as sunlight, land, water, labor etc.
are all available in the country. Agriculture is not just important for feeding the local population but to gain
foreign exchange. Sometime heavy rainfall can cause damage to farms and crops but if it is predicted earlier
than early warning can help to reduce the damage of life and resources. This paper provides a systematic
literature review of state of art machine learning and deep learning techniques proposed by various authors
to predict the rainfall. This paper gives information about Ensemble Learning, Logistic Regression, various
Linear Regression, Multiple Linear Regression, Artificial Neural Network, K-Nearest Neighbor, Support
Vector Regression, Decision tree and other miscellaneous models.