Flood Prediction and Recommendation System

Abstract

Floods are one of the most common natural disasters that occur, destroying human life and the environment. Flash floods can cause catastrophic damage to society. The impacts of the flood on society include loss of human life, worsening of health conditions due to waterborne diseases, loss of livestock, and many other factors. It could be difficult to set things back after a huge flood. The main aim of this work is to develop a flood prediction system by reading the rainfall that occurs and to predict the possibilities of floods as well as to recommend the nearby evacuation area for the people to move immediately. Thus, the idea of this chapter is to predict the occurrence of a flood using the rainfall data and recommend nearby evacuation areas away from flood hazard areas. Providing the utmost safety has been the major idea in working on this chapter. Perfect accuracy for rainfall data was obtained using the random forest algorithm. Five algorithms were chosen and applied to the training data set. Accuracy was found precisely while using random forest algorithm.