Artificial Intelligence Technique for Weather Parameter Forecasting

Abstract

The paper deals with the objective to study the different artificial intelligence methods and compare their efficiency of forecasting the temperature, rainfall, wind speed in order to contribute in policy making and forecast upcoming disaster if any. Daily data of weather parameters such as Minimum Temperature, Maximum Temperature, Relative Humidity, Evaporation, Bright sunshine, Rainfall, Wind Speed for Delhi region from January 1, 2017 to April 15, 2018 is considered. The behaviour of the considered data set is studied for weather parameters Temperature, Rainfall and Wind Speed daily basis and prediction are made and compared for the period April 16-30, 2018 using Multilayer perceptron (MLP), Radial Basis Function(RBF) and Sequential Minimal Optimization(SMO) artificial intelligence techniques. On comparing these methods, it is observed that MLP Regression shows the least error and maximum Correlation coefficient and is concluded to be the more efficient artificial intelligence technique for forecasting weather parameters. The study will help the concerned authorities for future planning and take preventive steps for the future coming calamities if any. It will also help the government to make effective policies.