An Ensemble Learning Based Model for Flood Prediction

  • C. Prabhavathi, S. Musfira, A. Nihitha, Y. Sindhuja Reddy

Abstract

Floods are the most devastating natural disasters which cause huge damage to person’s life. The research on the flood prediction tells that they contribute to risk reduction, decrease in agriculture loss, minimization in the loss of person life and reduction of property damage associated with it. In this paper we investigated various machine learning (ML) techniques for predicting floods. The main propaganda of this project is prediction of floods using ensemble learning and preventing those floods by sending alert signals. As a consequence, this paper introduces the best optimistic forecast models by ensemble learning for accurate prediction. Therefore, the significant techniques in improving the standard of flood prognosis methods are investigated. Here we use logistic regression and random forest algorithm for testing and training the dataset under supervised learning with ensemble technique. An app for this prediction and prevention of floods are created. This paper explains ensemble learning can often perform better than any single model. Therefore proposed system efficiency is measured by classification accuracy and user satisfaction with the model. An accuracy of 85% was achieved through this technique.

 Keywords: Classification, Ensemble Technique, Flood Forecasting, Flood Prediction, Flood Prevention, Logistic Regression, Machine Learning, Random Forest

Published
2020-05-28
Section
Articles