Flood predictions from metrics to classes by multiple machine learning algorithms coupling with clustering-deduced membership degree

Key Points

• Flood predictions significantly improve through the integration of multiple machine learning algorithms.
• Using clustering deduced membership degree enhances the accuracy of prediction models by 15% in various scenarios.
• Analysis of flood metrics using machine learning methods identifies crucial patterns that may inform better planning.
• This approach supports the need for more reliable forecasting tools, suggesting future research into real-time applications.