A synergistic approach: multi-purpose K-nearest neighbor and active learning Kriging for efficient failure probability function estimation

Key Points

• Efficient function estimation reduces computational costs in modeling failure probabilities, enhancing reliability.
• A notable reduction in failure probability estimation error was achieved through active learning techniques and k-nearest neighbor methods.
• Analysis utilizing a robust hybrid model combining k-nearest neighbor and active learning kriging techniques optimized the predictive accuracy.
• Results suggest that integrating these methods may enable more effective and efficient failure assessments in various applications.