cXGBoost adapts reduced-order models for engineering simulations

Researchers have developed a new framework called Constrained Extreme Gradient Boosting (cXGBoost) to improve the accuracy of reduced-order models (ROMs) used in engineering simulations. This method adapts the basis construction of ROMs by predicting Proper Orthogonal Decomposition (POD) bases as functions of system parameters. By leveraging geometric representations on the Grassmann manifold and imposing a norm constraint during training, cXGBoost ensures the validity of inverse mappings and preserves the geometric structure of predicted subspaces, demonstrating robust performance across various complex systems. AI

Summary written by gemini-2.5-flash-lite from 1 source. How we write summaries →

IMPACT Introduces a novel machine learning approach for enhancing the efficiency and accuracy of complex engineering simulations.

RANK_REASON This is a research paper detailing a new machine learning framework for adapting reduced-order models. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.LG →

• Constrained Extreme Gradient Boosting
• cXGBoost
• Proper Orthogonal Decomposition
• POD
• Grassmann manifold
• arXiv

• paper
• other