LSTM based flood prediction system

Abstract

Flooding due to natural calamities results in loss of life and property. It is necessary to save the precious human lives and other essential assets. Thus, flood forecasting is essential and it should be carried accurately. So, we are going to use Large Small-Term Memory [LSTM] neural network to forecast the flood with machine learning (ML) approach. LSTM takes River level, Rainfall data and water discharge as an input and give result as no flood or flood. LSTM is proved have accurate results. LSTM takes the decision based current input, previous output, knowledge data with it. As it has ability to remember the history and processing the large amount of data.its decisions are proved to be good wherever it is implemented.