Early Warning Flood Forecasting Using Long Short-Term Memory Network

Abstract

Abstract

Flooding is the natural disaster which leads to massive loss of life and property as well. India faces this situation every year and millions of people are being displaced due to loss of shelter. Early warning of flood disaster in corresponding locality provides sufficient time to protect their precious life and property. However, the range of flood prediction introduces the issue of cost, reliability and maintenance. We are proposing the solution to discussed problem and experimentation. The proposed system used Long Short-Term Memory (LSTM) network to design flood forecasting system which will give the prior warning of flood. LSTM network has accuracy in results and is cost effective.

Keywords: LSTM, flood forecasting, artificial neural network, recurrent neural network