Large-Scale and Expensive Optimization
Chair: Dra. Guadalupe Carmona Arroyo
Many real-world optimization problems in science, engineering, and industry involve large numbers of variables and computationally expensive function evaluations, such as high- fidelity simulations, physical experiments, or complex system designs. These problems pose significant challenges for traditional optimization methods due to their high dimensionality, nonlinearity, and limited evaluation budgets.
This special session aims to bring together recent advances in evolutionary computation, numerical optimization, and surrogate-assisted methods to address the scalability and efficiency issues in solving large-scale and expensive optimization problems. We invite contributions that combine theoretical insights, algorithmic innovations, and real-world applications.
Topics of interested include (but are not limited to):
- Decomposition-based methods and variable grouping strategies.
- Dimensionality reduction techniques for high-dimensional optimization.
- Large-scale evolutionary algorithms and coevolutionary approaches.
- Optimization under strict evaluation budgets.
- Surrogate-assisted optimization.
- Multi-objective and constrained optimization for expensive problems.
- Parallel and distributed implementations for scalability.
- Benchmarking and performance metrics for large-scale optimization.
- Techniques for handling dynamic and streaming graphs.
- Transfer learning and meta-model reuse in expensive environments.
- Real-world applications.
