Improving Performance of Ensemble Prediction Models for Mach Number in Wind Tunnels Using Metalearning

Accurate prediction of Mach numbers is of paramount importance in wind tunnel experiments. While ensemble prediction models have shown promising performance compared with single-model approaches, existing ensemble methods often prioritize ensemble generation while neglecting the critical aspect of ensemble fusion. This oversight has led to suboptimal precision in Mach number prediction. In this paper, we address this issue by proposing a novel approach that leverages metalearning for ensemble fusion. By incorporating metalearning techniques, our framework effectively combines the strengths of multiple base learners to enhance predictive accuracy. Extensive experiments conducted on a comprehensive data set demonstrate the superiority of our proposed approach over traditional ensemble methods. The results highlight the significance of ensemble fusion in achieving high-precision Mach number prediction in wind tunnel experiments. This research contributes to the advancement of ensemble modeling by emphasizing the importance of fusion techniques and provides valuable insights for improving ensemble prediction models in various domains.