A comparative machine learning framework for earthquake damage mapping in Turkey: incorporating MARS and ensemble models in the 2023 Kahramanmaraş earthquake using high-resolution Pléiades imagery

Key Points

• Damage mapping accuracy increases with ensemble models, highlighting their effectiveness in predicting earthquake damage.
• The analysis utilizes high-resolution Pléiades imagery collected during the Kahramanmaraş earthquake to assess damages accurately.
• Machine learning approaches, including MARS and ensemble models, enable better integration of data for mapping purposes.
• The findings indicate a significant potential for machine learning in improving disaster response strategies and resource allocation.