Researchers have developed a new method for creating surrogate models for complex transport problems, particularly those with thin boundary layers. This approach, termed Rationally Enriched Chebyshev (REC) trunk, combines Chebyshev polynomials with rational elements derived from the adaptive Antoulas-Anderson (AAA) algorithm. Evaluations showed that the REC-trunk DeepONet significantly improved accuracy in predicting scalar profiles, temperature, and concentration compared to vanilla and Chebyshev-trunk DeepONets, especially in challenging parameter ranges and while suppressing oscillations. AI
IMPACT This research could lead to more accurate and efficient AI models for simulating complex physical phenomena.
RANK_REASON Academic paper detailing a new method for surrogate modeling in machine learning. [lever_c_demoted from research: ic=1 ai=1.0]
Read on Hugging Face Daily Papers →
AI-generated summary · Google Gemini · from 2 sources. How we write summaries →
