On the performance analysis of rainfall prediction using mutual information with artificial neural network

<span lang="EN-US">Monsoon rainfall prediction over a small geographic region is indeed challenging task. This paper uses monthly means of climate variables, namely air temperature (AT), sea surface (SST), and level pressure (SLP) the globe, to predict seasonal summer monsoon state Maharashtra, India. Mutual information correlates from grid 10</span><span lang="EN-US">°</span><span lang="EN-US"> longitude X latitude with Maharashtra’s time series. Based on correlations, selected features respective longitudes are given as inputs an artificial neural network. It was observed that AT SLP could excellent accuracy. The performance test dataset evaluated through mean absolute error; root square error, correlation coefficient, Nash Sutcliffe model efficiency maximum capability individual variable for performed better in all evaluation parameters except capability, where combined 2 AT, SST predictors outperformed. SLP-only model’s comparable AT-only model. 1 found than 2.</span>