Preference learning based on maximizing membership degree with heterogeneous information for landslide early warning

Key Points

• The approach enhances early warning systems for landslides through optimized preference learning techniques.
• Findings reveal that maximizing membership degree leads to better classification in early warning scenarios.
• Analysis utilized heterogeneous information to improve predictive accuracy and achieve more reliable alerts.
• This study suggests that leveraging diverse data sources may enable more effective landslide risk management.