Summary of Bayesian Surrogate Training on Multiple Data Sources: a Hybrid Modeling Strategy, by Philipp Reiser et al.
Bayesian Surrogate Training on Multiple Data Sources: A Hybrid Modeling Strategy
by Philipp Reiser, Paul-Christian Bürkner, Anneli Guthke
First submitted to arxiv on: 16 Dec 2024
Categories
- Main: Machine Learning (stat.ML)
- Secondary: Machine Learning (cs.LG)
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 proposes novel probabilistic methods for integrating simulation data and real-world measurement data during surrogate model training. The authors highlight the limitations of using simulation models as approximations, citing issues with simplifications, misspecifications, and ignored hints from real-world data. They present two approaches: one trains separate surrogates for each data source, combining their predictive distributions; another trains a single surrogate incorporating both data sources. Case studies demonstrate improved predictive accuracy, coverage, and diagnosis of simulation model problems. This work has implications for system understanding and future model development. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Surrogate models help us solve complex problems by simplifying real-world systems. But these simplified models can be limited. New methods are needed to combine simulated data with real-world measurements to get a better picture. This paper presents two ways to do this: one uses separate surrogates for each type of data, while the other trains just one surrogate that combines both. The authors tested these approaches on synthetic and real-world examples and found they improve our predictions and help us identify problems in the models. |
