What does this research mean for the field?

The active-learning-driven deep neural network (ALDNN) meta model significantly reduces the number of limit state evaluations required for reliability analysis from 10^5–10^6 to approximately 10^2 while maintaining high accuracy. Novelty: ClaimNovelty.NOVEL_FINDING. Consensus alignment: ConsensusAlignment.NEUTRAL.

What question did this study set out to answer?

The aim is to develop a deep learning approach to enhance the reliability analysis of structural systems by reducing computational costs.

February 28, 2026MathematicsOpen Access

Active-Learning-Driven Deep Neural Network Meta Model for Scalable Reliability Analysis of Complex Structural and High-Dimensional Systems

View Full Paper

Ask AI

Bookmark

Authors

SLSangik Lee

Discussion

Member takes

Overview

Demonstrates a new algorithm that enhances reliability assessment in complex systems, suggesting improved efficiency and accuracy.

Key Points

The aim is to develop a deep learning approach to enhance the reliability analysis of structural systems by reducing computational costs.
Developed an active-learning-driven deep neural network (ALDNN) meta model algorithm.
Implemented a multi-phase active learning framework with weighted sampling.
Utilized adaptive threshold-based candidate filtering for model training.
Iteratively selected important points based on feedback from estimated responses.
Reduced the number of limit state function evaluations from 10^5–10^6 to approximately 10^2.
Maintained high accuracy in reliability assessments during benchmarking.
Verified applicability on complex frame structures using finite element methods.

Cite This Study

Sangik Lee (2026) studied this question.

synapsesocial.com/papers/69a287130a974eb0d3c026fe https://doi.org/10.3390/math14050796

View Full Paper

Ask AI

Bookmark

February 28, 2026MathematicsOpen Access

Active-Learning-Driven Deep Neural Network Meta Model for Scalable Reliability Analysis of Complex Structural and High-Dimensional Systems

View Full Paper

Ask AI

Bookmark

Authors

SLSangik Lee

Discussion

Member takes

Overview

Demonstrates a new algorithm that enhances reliability assessment in complex systems, suggesting improved efficiency and accuracy.

Key Points

The aim is to develop a deep learning approach to enhance the reliability analysis of structural systems by reducing computational costs.
Developed an active-learning-driven deep neural network (ALDNN) meta model algorithm.
Implemented a multi-phase active learning framework with weighted sampling.
Utilized adaptive threshold-based candidate filtering for model training.
Iteratively selected important points based on feedback from estimated responses.
Reduced the number of limit state function evaluations from 10^5–10^6 to approximately 10^2.
Maintained high accuracy in reliability assessments during benchmarking.
Verified applicability on complex frame structures using finite element methods.

Cite This Study

Sangik Lee (2026) studied this question.

synapsesocial.com/papers/69a287130a974eb0d3c026fe https://doi.org/10.3390/math14050796

View Full Paper

Ask AI

Bookmark