Summary of Conditional Pseudo-reversible Normalizing Flow For Surrogate Modeling in Quantifying Uncertainty Propagation, by Minglei Yang et al.
Conditional Pseudo-Reversible Normalizing Flow for Surrogate Modeling in Quantifying Uncertainty Propagation
by Minglei Yang, Pengjun Wang, Ming Fan, Dan Lu, Yanzhao Cao, Guannan Zhang
First submitted to arxiv on: 31 Mar 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Numerical Analysis (math.NA)
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 This paper introduces a novel approach to constructing surrogate models for physical systems contaminated by noise. The conditional pseudo-reversible normalizing flow model efficiently quantifies both forward and inverse uncertainty propagation without requiring knowledge of the noise or auxiliary sampling methods. By directly learning the conditional probability density functions, the trained model can generate samples from any function whose high-probability regions are covered by the training set. This approach simplifies implementation and enables theoretical analysis thanks to the pseudo-reversibility feature. The paper provides a rigorous convergence analysis using the Kullback-Leibler divergence and demonstrates its effectiveness on several benchmark tests and a real-world geologic carbon storage problem. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Surrogate modeling helps us understand complex physical systems that are noisy or uncertain. This paper creates a new way to build these models, called conditional pseudo-reversible normalizing flow. It can predict both how the system will behave in the future (forward) and what we would see if we went back in time (inverse). The best part is that it doesn’t need to know about the noise or use extra methods. This makes it easier to implement and understand. The paper shows that this approach works well on some examples and a real-world problem where carbon is stored underground. |
Keywords
* Artificial intelligence * Probability
