Summary of Multi-fidelity Residual Neural Processes For Scalable Surrogate Modeling, by Ruijia Niu et al.
Multi-Fidelity Residual Neural Processes for Scalable Surrogate Modeling
by Ruijia Niu, Dongxia Wu, Kai Kim, Yi-An Ma, Duncan Watson-Parris, Rose Yu
First submitted to arxiv on: 29 Feb 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: None
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary The proposed Multi-fidelity Residual Neural Processes (MFRNP) framework aims to improve multi-fidelity surrogate modeling by explicitly modeling the residual between aggregated lower fidelity data and ground truth at the highest fidelity. This approach optimizes lower fidelity decoders to capture both in-fidelity and cross-fidelity information, allowing for better inference performance, especially in out-of-distribution scenarios. MFRNP is shown to significantly outperform state-of-the-art methods in learning partial differential equations and a real-world climate modeling task. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Multi-fidelity surrogate modeling tries to create an accurate copy of the highest level data by combining lower-level information. Current methods struggle when dealing with big datasets. Some solutions use neural networks, but these approaches have limitations when it comes to making predictions in situations where there’s limited information from the highest level. To solve this problem, a new approach called Multi-fidelity Residual Neural Processes (MFRNP) was developed. MFRNP looks at the difference between what we know from lower levels and the actual data from the highest level, allowing for better prediction performance. |
Keywords
* Artificial intelligence * Inference
