.png){kind=icon}
.png){kind=icon}
Return
Dynamic multi-knowledge evolutionary algorithm for sparse large-scale multi-objective optimization
.png){kind=icon}
PRE
AI
DOI：10.1016/j.knosys.2025.114764.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
• Proposes a knowledge-guided evolutionary framework that integrates prior, filter, and statistical vectors to guide binary optimization in sparse large-scale problems adaptively. • Introduces a multi-interval sampling-based initialization strategy to estimate variable importance more reliably and support sparsity-aware population generation. • Enhances binary variation through two complementary operators, with the algorithm adaptively switching between early exploration and late-stage refinement. • Designs an adaptive real-valued variation operator with selective mutation and dynamic parameter adjustment for efficient fine-tuning of selected variables. • Demonstrates state-of-the-art performance on both benchmark and real-world sparse optimization tasks, including sparse neural network training and signal reconstruction.
Journal
K
IF:
7.6
Papers:
1.2W
Citations:
4.5W