An Enhanced Novel Data Visualization Based on Flood Prediction Method Using Highest and Lowest Rainfall Patterns by Comparing Decision Tree Over Logistic Regression

Accurate flood prediction is critical for effective disaster management and mitigation. This research introduces an innovative approach to data visualization for flood prediction that utilizes historical rainfall patterns. By analyzing the highest and lowest rainfall data, the study proposes a novel visualization technique that enhances the understanding of flood dynamics. We compare the effectiveness of decision tree algorithms against logistic regression in predicting flood likelihood. Our methodology centers on assessing these models’ capabilities to interpret complex rainfall data and predict potential flooding events accurately. A significant contribution of this work is the development of an intuitive visualization tool that can assist policymakers and emergency responders in making informed decisions. The results demonstrate the superior performance of decision tree models in terms of prediction accuracy and visualization efficacy. This study lays the groundwork for future research aimed at incorporating more sophisticated machine learning techniques into flood prediction models.