High-Dimensional Optimization

Memory-Guided Trust-Region Bayesian Optimization (MG-TuRBO) for High Dimensions

This research introduces Memory-Guided Trust-Region Bayesian Optimization (MG-TuRBO) to address high-dimensional, expensive optimization problems like traffic simulation calibration. It compares MG-TuRBO with other Bayesian optimization methods and genetic algorithms on real-world traffic scenarios.