Rainfall prediction based on 100 years of meteorological data
Sandeep Kumar Mohapatra, Anamika Upadhyay, Channabasava Gola · 2017
Weather being a random phenomenon its prediction has been always a challenge for the meteorologist all over the world. There are number of approaches for predicting this weather based on atmospheric data collected by various means. Our work focuses on use of data mining techniques for predicting rainfall of an area on basis of some dependent features like precipitation and wet day frequency. Instead of taking current data we are focusing on the data collected till date by the meteorological department. We are investigating a data mining technique using linear regression model on the data collected for wet day frequency, precipitation and rainfall for the years ranging from 1901 to 2002 of Bangalore, India. The regression model developed has been trained and validated against the actual rainfall of that area, which further was used to predict the rainfall in coming years. The performance of the algorithm was further boosted using Ensembles technique using K fold.