期刊名称：Journal of Theoretical and Applied Information Technology
印刷版ISSN：1992-8645
电子版ISSN：1817-3195
出版年度：2021
卷号：99
期号：19
语种：English
出版社：Journal of Theoretical and Applied
摘要：The flood is the cause of destruction for many places in the world. Flood prediction is a complex method due to its nature. The flood arrives with vast destruction in the society. The flood assessment is an essential task for the government bodies to take measures at the right time. In this study, a new procedure is designed to choose the best machine learning model for efficient flood mapping using machine learning techniques. An automated algorithm is created using a Decision Tree, Random Forest, Gradient Boosting Classifier, and KNN techniques. The Random Forest Classifier gave the highest accuracy with 88.58% with this (the considered) dataset. For the evaluation of the model, confusion matrix, learning curve, and classification reports are used. In this study, Open Flood Risk by postcode dataset is used.
